P E R S P E C T I V E S

What role - if any - shouldf incenfives play in the v^orlcplace?

Rethinkinq Rewards
It is difHcult to overstate the ex-
tent to which most managers-and
the people who advise t hem- be-
lieve in the redemptive power of
rewards, AHie Kohn argues in "Why
l nccnt i vf Plans Cannot Work"
(September-October 1993). Certain-
ly, the vast majority of U.S. corpora-
tions usf some sort of proj^ram in-
tended to motivate employees by
tyin^ compensation to one index of
performance or another. But more
striking is the rarely examined belief
that people will do a better job if
they have been promised some sort
of incentive.
This assumption and the practices
associated with it are pervasive, but
a Hfiwinj; collection of evidence
snpports an opposing view. Accord-
ing to numerous studies in labora-
tories, workplaces, classrooms, and
other settings, rewards typically nn-
dermine the very processes they are
intended to enhance. In Kohn' s
view, the findings snggest that the
failure of any given incentive pro-
gram is dne less to a glitch in that
program than to the inadequacy of
the psychological assumptions that
ground all such plans.
Do rewards work? The answer de-
pends on what we mean by "work."
Research suggests that, by and large,
rewards succeed at securing one
thing only: temporary compliance.
They do not create an cndnring com-
mitment to any value or action.
They merely, and t emporari l y,
change what we do. According to
Kohn, incentives in the workplace
simply can't work.
Nine experts consider the role of
rewards in the workplace.
G. Bennett Stewart III
Senior Partner
Stern Stewart & Co.
New York, New York
A world without A's, praise, gold
stars, or incentives? No thank you,
Mr. Kohn. Communism was tried,
and it didn't work.
The Soviet and Chinese econo-
mies collapsed because people were
not allowed to share in the fruits of
their individual efforts. With gains
from personal initiative harvested as
a public good, innovation ceased.
PHOTO BY TONY RINALDO 37
P E R S P E C T I V E S
and productivity froze. "They pre-
tend to pay us, and we pretend to
work" was the Russian worker' s
lament for the system Kohn now
proposes. But for pay to mean any-
thing, it must be linked to per-
formance. Without that link, pay
becomes nothing more than enti-
tlement, a joh nothing more than
a sinecure.
Kohn is unhappy that rewarding
some people necessitates penaliz-
ing others. Winston Churchill's apt
aphorism is the best response. He
said, "The virtue of communism is
the equal sharing of its misery, and
the vice of capitalism is the unequal
sharing of its blessings." You can't
have it both ways, Mr. Kohn, You
simply can' t have the equality of
outcome you desire with the robust,
dynamic economy we all want.
Contrary to the small-sample psy-
chology tests Kohn cites, the respon-
siveness of ordinary citizens to in-
centives is demonstrated daily in
our economy. Consumers cut con-
sumption in reaction to the "penal-
ty" of a price increase and raise pur-
chases in reaction to the "hrihe" of
a lower price. The price system effi-
ciently allocates scarce resources
precisely because it rewards people
who conserve and penalizes those
who fail to respond. Can it be true,
as Kohn seems to think, that people
respond to monetary incentives
when they spend their income but
not when they earn it;
If Kohn makes a useful point, it is
when he says that people won't want
to he paid for doing specific tasks.
But here is where we disagree: peo-
ple should he rewarded for an overall
joh done well. To put the point in
economic terms, the best incentive
is having a piece of the action. Com-
pany stock, however, is not the best
approach to instilling ownership, for
it frequently leaves too loose a link
between pay and performance.
The best approach often is to carve
employees into a share of the profit
contributed by their part of the com-
pany. Profit should be defined in rel-
evant cash-flow terms after covering
the cost of all capital employed, a
measure that Stern Stewart & Co.
calls Economic Value Added. EVA
provides employees with three clear
38
incentives: tu unpiovc piotitahility,
to grow profitability, and to with-
draw resources from uneconomic ac-
tivities. In addition, it ties their deci-
sions and energies directly to the
"net present value" of their enter-
prise. All key managers at Quaker
Oats have been on an EVA sharing
plan for several years, and Scott
Paper Company introduced an EVA
incentive program for all salaried
employees at the beginning of 1993,
to name but 2 of the 50 prominent
companies that have adopted this
approach in recent years.
Eileen Appelbaum
Associate Research Director
Economic Policy Institute
Washington, D.C.
Companies today are under in-
tense pressure to improve efficiency
"A world
without A's,
praise, gold
stars, or
incentives?
No thank you,
Mr. Kohn.
Communism
was tried,
and it didn't
work/'
G. Bennett Stewart III
and quality at a time when their re-
sources are severely limited. Fid-
dling with compensation schemes
appeals to many managers as a cheap
way to improve their companies'
performance hy providing individu-
als with incentives to work harder.
In fact, reliance on individual incen-
tives to motivate workers and spur
productivity has a long history in
the United States. The U.S. human-
resource model evolved in the 1950s
partly in response to then-current
theories of industrial psychology. By
designing compensation schemes
that recognize and reward individual
differences, companies expected to
reap the rewards of increased em-
ployee motivation and improved job
performance. This idea continues to
inform present managerial thinking.
In his article "Why hicentivc Plans
Cannot Work," Alfie Kohn has per-
formed an important service by mar-
shaling the modern evidence on the
psychological effects of incentives
and by showing that rewards fail
to improve, and may even reduce,
performance.
We arc still left, however, with
questions ahout what improves a
company's performance and what
role compensation actually plays in
that improvement. I would offer the
following answers, based on an anal-
ysis of nearly 200 academic case
studies and consultants' reports, car-
ried out with Rosemary Batt - a doc-
toral candidate in labor relations and
human-resource policy at MIT's
Sloan School of Management-and
published in The New American
Workplace, forthcoming from the
ILR Press in 1994.
In the early part of the twentieth
century, workplace innovations at-
tempted to improve employee satis-
faction and, at the same time, com-
pany performance. In contrast, the
move to high-performance work sys-
tems since the mid-1980s is moti-
vated by the need to improve quality
and reduce costs simultaneously. In
the mass-production model of work
organization, whether the Taylorist
or the U.S. HR version, improving
quality raises costs-for inspection,
supervision, rework, and waste. It
was quite a shock to U.S. sensibili-
ties, therefore, when Japanese auto
P E R S P E C T I V E S
manut';icturers demonstrated that
new ways of organizing work could
delivernoticeahly higher quality and
customer satisfaction at signifi-
cantly lower prices. It took nearly a
decade for companies in the United
States to realize that they would
h;ivc to change.
Our review of the evidence indi-
cates an acceleration of experimen-
tation with innovative workplace
practices and the emergence since
the mid-1980s of two distinctly
American high-performance mod-
els: a U.S. version of lean production
that relies on employee involvement
and a U.S. version of team produc-
tion that relies on employee empow-
erment for performance gains. Pro-
Lluctivity and performance improve
the most when work is reorganized
so that employees have the training,
opportunity, and authority to partic-
ipate effectively in decision making;
wlien they have assurances that they
will not be punished for expressing
unpopular ideas; when they realize
that they will not lose their jobs as a
result of contributing their knowl-
edge to improve productivity; and
when they know that they will re-
ceive a fair share of any performance
gains, assurances which unionized
workers in high-performance com-
panies cnioy.
Attempts to improve performance
by mani pul at i ng compensat i on
packages have proven counterpro-
ductive. However, reorganizing the
work process to capitalize on em-
ployee skills and participation has
improved performance, especially in
combination with employment se-
curiiy, gaiiishiiriiig, and incentives
to take part in training. In this sense,
then, compensation packages are an
important component of the human-
restiurcc practices that are neces-
sary to support high-performance
work systems.
vatitm and organizational effective-
ness. But because certain practical
considerations and cultural differ-
ences arc not addressed, the argu-
ment is flawed.
Like Kohn, I have found that many
managers in the United States and
the United Kingdom-hut not, inci-
dentally, in continental Europe or
Japan-have deeply held assump-
tions about the role of incentive pay
in motivation. These assumptions
lead them to engage compensation
consultants in answering the wrong
question: How should we design the
incentive system in order to obtain
the desired behavior? The more im-
portant question is: What role, if
any, should incentive compensation
play? Like Kohn, I have found that
assumptions about incentive com-
pensation have led many managers
to expect incentives to solve organi-
Michael Beet
Professor of Business
Administration
Harvard Business School
Boston, Massachusetts
Kohn has mounted an eloquent ar-
gument, when it is considered in
light of what we know about moti-
''If incentive
systems do not
motivate, what
should
managers do
ahout
compensation?
Surely, Kohn
would not
suggest that
everyone
should he paid
the same."
Michael Beer
zational pruhlems, when there are
actually deeper underlying reasons
for those problems.
Managers tend to use compensa-
tion as a crutch. After all, it is far
easier to design an incentive system
that will do management' s work
than it is to articulate a direction
persuasively, develop agreement
about goals and problems, and con-
front difficulties when they arise,
The half-life of an incentive system
is at best five years. When it stops
paying off, employees turn against
it. And the result is another dysfunc-
tional by-product of incentive sys-
tems: precious attention, time, and
money is expended on endless de-
hates about and redesigns of the in-
centive system.
If incentive systems do not mo-
tivate, what should managers do
about compensation? Surely, Kohn
would not suggest that everyone
should be paid the same. In some
industries or functions-sales, for
example-incentive compensation
is the prevailing practice. In these
areas, without paying for perfor-
mance, an organization will lose its
best people. Yet by paying for perfor-
mance, the company rims the dan-
ger of encouraging self interest in-
stead of organizational commitment.
This is a fundamental pay-for-per-
formance dilemma that practicing
managers confront and that Kohn
neglects to address.
It is undoubtedly true that in to-
day's competitive environment, in-
terdependence between different
business units and functions as well
as the need for custtimcr service and
quality make incentive compensa-
tion less appropriate than it once
was. But there are circumstances in
which it is the only solution avail-
able: for example, managers of inde-
pendent stores far from headquar-
ters who don't have a motivating
manager-subordinate relationship or
salespeople whose performance is
independent of other business units
and who operate without supervision
much of the time.
Managers who agree with Kohn
should pay for performance but
strive to use incentive systems as lit-
tle as possible. Pay is an exercise in
continued on page 42
39
smoke and mirrors. Companies can-
not stop paying for performance.
However, they should avoid using
incentives for all the reasons that
Kohn suggests.
What can managers do? They
should focus on paying people equi-
tably, rather than using pay as an
instrument of motivation. They
should avoid coupling pay with year-
ly or quarterly performance, while
promoting the top 10% or 15% of
employees for outstanding long-
term contributions. The poorest
performers should be weeded out,
while the rest should be praised for
good performance and recognized
through other means to promote
self-esteem.
We are indebted to Kohn for ring-
ing the alarm, but he docs not
provide managers with creative,
practical solutions to the pay-for-
performance dilemma.
Andrew M. Lebby
Senior Partner
The Performance Group
Washington, D.C.
The effect of rewards on motiva-
tion and performance is one of the
most studied subjects in the man-
agement literature. Year after year
we validate the finding that employ-
ees' perceptions of underpay result
in decreased productivity, while in-
creased pay doesn't result in in-
creased productivity. Year after year
we ask employees what motivates
them, and year after year they reply:
a sense of accomplishment in per-
forming the work itself, recognition
from peers and top management, ca-
reer advancement, management
support, and, only then, salary.
If Kohn is unable to find data that
support anything but a negative rela-
tionship between finaneial incen-
tives and performance, why is it that
in the face of overwhelming evi-
dence executives continue to hold
onto ineffective methods; Why is it
that they refuse to provide those
things that employees say tbey
want, that direetly relate to in-
creased productivity, and that have
little or no financial cost?
When we stop to separate the
physical nature of the reward it-
i
''Intrinsic
LOtivation-
being
motivated by
cballenge and
enjoyment-is
essential to
creativity. But
extrinsic
motivation-
being
motivated by
recognition and
money-doesn' t
necessarily
burt/'
Teresa M. Amabile
self from what the recipient finds
rewarding, some possible answers
appear. When we ask employees,
"What was tbe last reward you re-
ceived?" the most frequent response
is some variant of "money." When
we ask, "Wbat did you find reward-
ing about money?" the most fre-
quent response is that it was a tacit
acknowledgment of the outstanding
nature of their contribution. Just as
it is easier for some parents to sbow
love witb gifts tban with hugs, it is
often easier for organizations and
managers to sbow gratitude with
money than with words.
Our current notions of pay follow
naturally from our antiquated, Tay-
lorist, mechanistic models for de-
signing work. The work we do and
how we do it have shifted signifi-
cantly, but our reward and salary
structures remain essentially the
same. Senior managers will end ti-
nancial incentives only when they
rethink what work is and how it is
performed. Organizations that have
redesigned work to reflect cross-
functional business processes or
tbose tbat bave implemented the
actual principles of TQM bave had
to rethink pay and performance.
Employees have said, "Give us the
tools, tbe skills, the information, the
support, and the respect we need."
In different words, "Give us real
capital, intellectual capital, and
symbolic capital, and we'll increase
your-and our-financial capital."
Money is an outcome of bigb per-
formance. Satisfaction and respect
are incentives to it.
Teresa M. Amabile
Professor of Psychology
Brandeis University
Waltbam, Massachusetts
Kohn is absolutely right when he
tells us that rewards can work
against real commitment and cre-
ativity. But he doesn't tell the whole
story. There are important differ-
ences between bribes and equitable
compensation, and tbcrc are condi-
tions under wbieb rewards can in-
crease involvement and creativity.
What matters is what those rewards
actually mean.
As Kohn points out, there is abun-
dant evidence tbat interest and per-
formance decline over the long run
when people feel they arc controlled
by incentive systems or any otber
management system. What Kohn
fails to point out is that people do
not always feel controlled by re-
wards. In a reeent study of profes-
sional artists, my students and I
found, as Kohn would bave predict-
ed, that noncommissioned works
were more creative tban commis-
sioned works. However, what mat-
tered was not the obvious fact of
contracting for reward, but tbe de-
gree to which tbe artist felt con-
strained by the terms of the commis-
sion; tbe more constraints, the lower
the creativity. In fact, some artists
considered some of their commis-
sions enabling, allowing tbem to
create an interesting work of art tbat
they wouldn't otherwise have had
42
HARVARD BUSINESS REVIEW November-December 1993
the mciins tu do. When the reward
presented the artist with new possi-
bilities, in other words, creativity
actually increased.
Intrinsic motivation-being moti-
vated by ehallenj;c and enioyment-
is essential to creativity. But extrin-
sic motivation-heing motivated by
recognition and money-doesn't nee-
essarily hurt. The most creative
artists in our study tended to be mo-
tivated more by challenge, but they
also tended to be mot i vat ed by
recognition, Kohn accurately docu-
ments the evidence that rewards can
undermine creativity. But he fails to
mention the evidence tbat tangible
rewards can actually enhance cre-
ativity under certain circumstances,
most notably when the individual's
primary focus is on the intrinsic re-
ward t)f the work itself.
Bribes, as Kohn frequently notes,
are bound to make people feel con-
trolled, and he rightly points out
their negative effect on people's
work. But he implicitly includes
salary in the same category as bribes
when he argues that "pay is not
a motivator." Certainly, there are
some circumstances under which
salary increases arc perceived as
bribes. A few years ago, for example,
I interviewed an R&D scientist who
was widely considered to be one of
the three most important innovators
in a targe, suecessful company; he
was also considered extremely ec-
centric, "They offered me a pretty
large salary increase this year, but
I refused it," he recounted. "Right
now, my lab is my playground; I
pretty much come in here and do
things the way I want. But the more
they pay you, the more they think
they own you."
A much more common reaction,
however, was the feelinj; expressed
by other scientists that their salary
increases recognized their creative
contributions. Generous compensa-
tion, including companywide profit
sharing, need not be seen as a bribe,
particularly when it is presented as
the equitable outcome of creative
competence.
Although Kohn's article is clear
about what managers should avoid,
it has little to .say about alternatives
to incentives. There is much that
''Appropriate
rewards for
improved
performance
have always
made good
sense,
intuitively and
practically.
They aren't
wrong.
They aren't
intrinsically
demotivating."
Jerry McAdanis
can be said about redesigning work
and the work environment so that
extrinsie motivators become less
central. Managers need to know how
to use these alternative techniques
before they can be expected to aban-
don the incentive systems on which
they have relied for so long.
If Kohn can convince even a few
managers that incentive plans are
not the keys to innovative, high-
quality performance, he will have
made a significant contribution. But
it would be a mistake to believe that
reward and recognition must always
have a negative effect on perfor-
mance or that creative people can-
not be motivated by both money and
interest in the work itself. As the
poet Anne Sexton once said, "I am in
love with money, so don't be mistak-
eti. But first I want to write good
poems. After that, I am anxious as
hull to earn money and tame and
bring the stars all down."
Jerry McAdams
Vice President, Performance
Improvement Resources
Maritz Inc.
Director
Consortium for Alternative Reward
Strategies Research
St. Louis, Missouri
A few years ago, Kohn did the
business community a service with
his book, No Contest: The Case
Against Competition, which argues
that competition is for the market-
place rather than the workplace.
The book makes a compelling ar-
gument for focusing on teamwork
instead of pitting one employee
against another. The key to success,
Kohn maintains, is to create an at-
mosphere of cooperation, channel-
ing employees' creativity and energy
to affect the business objectives of
the organization positively. Compe-
tition between individuals, on the
other hand, only gets in the way.
Now Kohn argues that rewards get
in the way as well. On the basis of
my 20 years of researching; and de-
signing reward plans for sales and
nonsales employees, I disagree. Ap-
propriate rewards for improved per-
formance have always made good
sense, intuitively and practically.
They aren't wrong. They aren't in-
trinsically demotivating. Data show
they make good business sense.
Of course, there is always a mar-
ket for speeches, books, and articles
that profess, through highly selec-
tive academic research, that what is
working really i.'iu 't. Kohn's article is
a provocative exercise in attention-
getting, niche marketing. Unfortu-
nately, Kohn's article will probably
be used by some to deny perfor-
mance-improvement opportunities.
I do agree with Kohn's point re-
garding the negative aspects of the
reinforcement of tasks, particularly
when the reinforcement plan is
piece-rate or merit-pay based. Mea-
suring and rewarding on an individu-
al level (sales excepted) does tend
to become controlling. Tbe focus
should he on business objectives,
not tasks. The study, Capitalizing
HARVARD BUSINESS REVIEW November-December 1993 43
P E R S P E C T I V E S
on Human Assets, covering one-mil-
lion employees and 432 compensa-
tion plans and sponsored by the non-
profit Consortium for Alternative
Reward Strategies Research (CARS),
shows that rewarding groups of em-
ployees, usually wbole plants and of-
fices, is a powerful business strategy.
According to the study, this strat-
egy pays off a median three-to-one
return on the cost of the rewards.
Employees earn from 2% to 15% of
their base pay in incentives or non-
cash awards. No layoffs appear to
result from the improved perfor-
mance. Interviews and extensive da-
ta analysis of the 432 plans show
positive employee-management co-
operation and improved information
sharing and employee involvement.
Rewards are not bribes. Bribes are
payments for behavior that may be
in the organization's best interest
but are clearly not in the individ-
ual's. Rewards reinforce a "win-
win" environment. The ohjective of
a reward plan is not to "control or
manipulate," as Kohn contends. It is
to provide focus and reward im-
proved performance.
Tom Peters was right when he
wrote about Kohn's thesis, "What
we need is a lot more positive rein-
forcement, and a lot less of the nega-
tive kind, throughout the corporate
landscape. And far from cautioning
companies about the dangers of in-
centives, we should be applauding
those that offer their employees a
bigger piece of the action" (INC,
April 1988). The CARS research has
done just that, looking at more plans
in greater depth than any other
study. The bottom line is simple:
reward plans work when properly de-
signed and supported; there can be
something in it for everyone,
I think it is time to focus on the
productive use of people as assets to
business not on the counterproduc-
tive theories in Kohn's article.
L. Dennis Kozlowski
Chairman and CEO
Tyco Laboratories, Inc,
Exeter, New Hampshire
I'll accept that elephants cannot
fly and that fish cannot walk, but
Kohn's argument that incentive
plans cannot work ciuries the laws of
nature at Tyco Laboratories. Tyco
provides a compelling case study
that incentives can and do work for
both managers and shareholders. In
fact, we believe our incentive com-
pensation program is at the heart of
our company's success.
We view the relationship between
Tyco's management and its share-
holders as very straightforward:
management works for the share-
holders. It is our mission to create
value for them through stock-price
appreciation. In fact, our share price
has closely tracked our earnings
curve for many years, lending con-
siderable weight to our determina-
tion to encourage earnings growth in
a prudent and consistent manner.
Our compensation program, in turn,
was designed to align the financial
interests of our executives with
'TU accept that
elephants
cannot fly and
that fish cannot
walk, hut
Kohn's
argument that
incentive plans
cannot work
defies the laws
of nature at
Tyco
Lahoratories/'
L. Dennis Kozlowski
those of our shareholders. The basic
rule is this: the more the executives
earn for the shareholders, the more
they earn for themselves.
Tyco's 250 profit centers fall into
four major businesses. Within the
context of a few corporate financial
controls, we tell each profit-center
manager to run the business as if he
or she owned it. A decentralized ap-
proach lets us put the financial re-
sources of a $3-billion corporation
behind the entrepreneurial spirit,
drive, and resourcefulness of man-
agers who think and act like owners.
It's the best of both worlds. Profit-
center autonomy and responsibility
go hand in hand. We encourage each
unit's management team to share
the unit's profits. The more profits
the business unit earns for the share-
holders, the more compensation the
management team earns for itself.
Our incentive plan has several im-
portant and unique features. For one,
incentive compensation is directly
tied to each business unit's perfor-
mance and not to corporate results
or other factors beyond any individ-
ual's control. In addition, the awards
are not based on bow units perform
against a budget or any other preset
goal. Instead, awards constitute a
preestablished percentage of earn-
ings. Since we adopted this ap-
proach, the quality of the budgeting
process has substantially improved.
Finally, award opportunities are un-
capped, and, as a result, they encour-
age the entrepreneurial spirit that
we value.
When designed effectively and
integrated thoroughly into the man-
agement process, executive incen-
tive programs work well for manage-
ment and shareholders alike.
George F. Baker III
Associate Professor
Harvard Business School
Boston, Massachusetts
The problem is not that incentives
can't work but that they work all too
well. Kohn's analysis of the unin-
tended and unwanted side effects of
many incentive plans is perfectly
apt; plans that provide incentives for
the wrong behavior will produce the
wrong results. However, Kohn's so-
44
P E R S P E C T I V E S
lution to abandon iticcntivc plans
entirely is misguided. Rather, man-
agers must learn how to harness and
use the power of ineentives to drive
individual motivation and organiza-
tional effectiveness.
In several places, Kohn's asser-
tions about the weakness of incen-
tive plans only serve to highlight the
power of such plans to influence he-
havior. What Kohn says is absolute-
ly true: if teamwork and cooperation
are desired, and the incentive plan
rewards only individual results, then
the plan will generate counterpro-
ductive results. However, a well-de-
signed ineentive plan that rewards
team prodtietivity not only will
avoid such unproductive behavior
hut also will induce employee coop-
eration. This is the logical basis for
the majority of profit-sharing and
employee stock-ownership plans,
whose effectiveness mounting evi-
dence supports.
Similarly, Kohn's ohservation that
incentive plans cause employees to
curry favor with the hoss and with-
hold information ahout poor perfor-
mance is often accurate. But the so-
lution is not to eliminate the boss's
ability to reward employees. Instead,
supervisors should he trained to ig-
nore or punish politicking. It is pre-
cisely hecause incentives are so
powerful that Kohn can predict that
if managers reward politicking, poli-
ticking will result.
Reward plans need not be control-
ling, as Kohn seems to imply. Con-
sider the store-manager ineentive
plan at Au Bon Pain. Store managers
are given a profitability target and
are allowed to keep a substantial
fraction of any profits they earn
above this target. The chain puts few
constraints on how they achieve or
exceed their targets. The plan has
hardly been "the enemy of explo-
ration." Rather, it has resulted in an
explosion of entrepreneurial experi-
mentation and innovation. Notice,
h<jwever, that the Au Bon Pain plan
is not, in Kuhn's words, "contingent
on behavior." It is contingent on re-
sults, and herein lies the crucial dif-
ference. Plans that are contingent on
behavior will encourage the pre-
scribed hehavior and stifle initia-
tion. However, plans that reward de-
sired results arc likely to stimulate
innovation.
Perhaps the most disturbing omis-
sion from Kohn's article is his failure
to suggest an alternative to the use
of incentive plans. If companies are
to abandon extrinsic incentives as a
way to motivate employees, what
are they to use instead; Is Kohn rec-
ommending that we live with the
loss of individual motivation and
lack of organizational innovation
and flexibility that characterizes
companies and societies without ex-
trinsic incentives? Without some
level of extrinsic incentive to sup-
plement the intrinsic drive of indi-
viduals, organizations beeome un-
wieldy and inflexible, As a general
prescription for the management of
organizations, Kohn's approach is
naive and Utopian. In the real world,
organizations must manage incen-
tives if they arc to be fluxibli;, inno-
vative, and directed.
"Incentives are
neither all good
nor all bad.
Although not
the right
answer in all
cases, they can
he highly
effective
motivational
tools/'
Donita S. Wolters
Donita S. Wolters
Manager of Human Resources
IMM Operational Serviees, Inc.
Denver, Colorado
While Kohn makes a number of
valid points with respect to the dan-
gers of ineentive plans, his summary
execution of incentives is unwar-
ranted. Incentives are neither all
good nor all had. Although not the
right answer in all cases, they can he
highly effective motivational tools
and should be employed under the
appropriate circumstances.
Without a doubt, finaneial re-
wards can be, and have been, both
overused and misused. Implement-
ing a poorly designed or ill-suited
incentive plan can do more harm
than good because employees will
inevitahly receive mixed, even con-
flicting, messages from the organiza-
tion about its values and priorities,
leading to confusion and frustration.
Incentives are no substitute for good
management and should not be used
indiscriminately to remedy prob-
lems when more effective solutions
exist. Kohn mentions training and
goal setting as examples of effec-
tive strategies for improving produc-
tivity, and his advice is well-taken.
Incentives cannot improve perfor-
mance if employees are not properly
trained to perform their tasks or
have no idea what is expected of
them. But something more is of-
ten needed to elicit the necessary
effort. The ioh-rate pay systems that
typify unionized hlue-eollar environ-
ments-where mediocrity and lack
of innovation are the hallmarks,
and employees do just enough to
get hy - illustrate the point.
I have observed, as a veteran of
many employee-eounseling ses-
sions, that employees are more apt
to become disillusioned with incen-
tive plans when they feel exploited
because the expected rewards are
not forthcoming, not when they are
rewarded for something they were
inclined to do in the first place. To
avoid pereeptions of exploitation
and manipulation, however, two de-
continued on pa^e 48
45
P E R S P E C T I V E S
sign features of the incentive pro-
gram are imperative.
First, the criteria for and the actual
evaluation of performance must be
seen as ohiectivc and within the per-
former's control. This means that
anyone should be able to predict the
reward consistently and reliably
based on given actions and results.
The reward should not he deter-
mined through highly subjective
processes, such as a supervisor's in-
dividual opinion. Kohn seems to
support this view when he states
that "not receiving a reward one had
expected to receive is...indistin-
guishable from heing punished."
Second, the recipient should con-
sider the reward equal to the effort
that produced it. Too insignificant
and the incentive will be insulting
and thus ineffective; overdone and
the balance of fairness will be upset.
Insufficient attention to these dy-
namics may underlie the apparent
failure of many executive incentive
plans, which could more accurately
be termed entitlement programs.
Kohn goes on to decry the inabili-
ty of incentives to "create an endur-
ing commitment to any value or ac-
tion." I question the relevance of
this criticism. The purpose of incen-
Alfie Kohn Responds:
The average U.S. company has
come to resemble a game show:
"Tell our employees about the
fabulous prizes we have for them
if their productivity improves!"
None of my respondents doubts
the pervasiveness of this men-
tality. In fact, several profess
incredulity that anyone would
question the value of dangling
rewards in front of people. In my
experience, this reaction most
often comes from the consultants
who make their living selling
incentive programs. What I hear
aroLind the country from people
with no axe to grind is a frank
acknowledgment that incentive
plans rarely work.
Consider the following:
D A human-resource executive at
a major U.S. auto company re-
cently surveyed her colleagues in
various industries; they told her
that, at best, their incentive plans
didn't do too much damage.
D Training Magcizine ran a cover
story in August: "Why No One
Likes Your Incentive Program."
D As Michael Beer observes, pay-
for-performancc programs are
typically tossed out a few years
after they are begun.
D To the hest of my knowledge,
no controlled study has ever
found long-term improvement in
the quality of performance as a
result of extrinsic rewards.
Of course, it is comforting to
believe that incentives fail on-
ly for incidental reasons, such
as that they are "misused," as
Donita Woltcrs would have it, or
that they are offered "for the
wrong behavior," as George Baker
claims. But 1 helievc incentive
plans must fail, hecause they are
hased on a patently inadequate
theory of motivation. Trying to
undo the damage hy adopting a
new pay-for-performance scheme
is rather Uke trying to cure alco-
holism by switching from vodka
to gin. This argument makes a lot
of people angry, as seems clear
from Jerry McAdam's unpleasant
speculations about my ulterior
for-performance in particular.
Neither can produce quality, but
only the latter is positively harm-
ful. I agree with Aniahilc that
"generous compensation...need
not be seen as a hrihe," but I dis-
agree that "people do not always
feel controlled hy rewards."
Richard Ryan and his colleagues
at the University of Rochester, pi-
oneers in researching this ques-
tion, have concluded that "re-
wards in general appear to have a
controlling significance to some
extent and thus in general run the
motives and from the amusing, if
predictable, mutterings about
communism hy G. Bennett Stew-
art. If the attachment to carrot-
and-stick psychology - or any
dogma-is deep enough, ques-
tioning simply isn't permitted.
W. Edwards Deming, and oth-
ers hefore him, have been telling
us for years that money is not o
motivator. Judging from Teresa
Amabile's response, however, I
may not have heen clear enough
ahout the difference between
compensation in general and pay-
risk of undermining intrinsic mo-
tivation." Offering good things to
people on the contlition that they
do what you tell them is, almost
hy definition, a way of trying to
exert control.
But even someone who insists
that it's possible in theory to de-
vise a noncontrolling reward has
to concede that control is what
incentive plans in the real world
are all about. Just listen to the de-
fenders of these programs: the
whole idea is to "direct [employ-
ees'] behavior," as Wolters says.
48 DRAWING BY MARK STEELE
tives is not to change employees'
values but to direct their behavior in
ways that will henetit the organiza-
tion and the employees themselves.
More telling is Kohn's failure to
identity a viahle alternative to in-
centives. Of course, the intrinsic re-
wards he praises are extremely moti-
vating where they happen to exist,
but they are not always present and
cannot usually be created,
The current trend in organizations
is toward less hierarchy and more
t eamwork. For employees, this
means that fewer promotions are
available and greater eooperation
among coworkers is required. For
employers, this means that maxi-
mum versatility and productivity
must be summoned trom all mem-
bers. The use of incentive plans rep-
resents one strategy for aligning or-
ganizational and individual goals by
treating employees as partners in
both the risks and the successes of
the business. Kohn recognizes that
the majority of companies in the
United States utilize some sort of in-
centive plan. Indeed, his assertions
are being tested on the firing line and
disproved by a persuasive cross sec-
tion of U.S. business. _ ^
Reprint 93610
1 believe incentive plans must fail.
No wonder the evidence shows
that incentives do not "supple-
ment the intrinsic drive of indi-
viduals," as Baker believes, but
tend to supplant it. As a rule, the
more salient the extrinsic moti-
vator, the more intrinsic motiva-
tion evaporates.
One could say, as Baker does,
that incentives work too well, in
the sense that they are destruc-
tive of excellence and interest.
But one cannot conclude from
this that the problem is merely
one of implementation. Baker
behavior. Rather, the use of re-
wards and the extrinsic orienta-
tion tbey produce inexorably lead
people to focus on pleasing those
in charge of handing out the good-
ies. Fine-tuning the incentive
plan cannot solve the problem.
Finally, a number of correspon-
dents are understandably curious
about my views on what should
replace incentive plans. If a dis-
cussion on this point was con-
spicuously absent from the arti-
cle, wbich was an excerpt from
my book Punished hy Rewards, it
errs in assuming that just because
rewards undermine cooperation
it follows that tbey can also cre-
ate it. If something has the power
to hurt, that doesn't mean more
of it will motivate. Again, think
of money: less of it can demoti-
vatc, but that doesn't mean that
more of it will motivate. I think
Baker also misunderstands why
employees try so hard to con-
vince their reward-dispensing su-
pervisors that everything is under
control. It's not because the latter
are deliberately rewarding such
was due to limited space. I do
grapple at length with alterna-
tives to incentives in another
chapter, "Thank God It's Mon-
day." Here, a few words will have
to suffice.
On compensation, my advice is
this: pay people well and fairly,
then do everything possible to
help tbem forget about money. I
bave no objection to profit-shar-
ing: it seems sensible enougb that
the people who made the profit
ought to have it. Nor am I keen to
promote one criterion for com-
pensation over another: for exam-
ple, need, seniority, job responsi-
bilities, training, market value.
My concern is primarily to con-
vince managers to stop manipu-
lating employees witb rewards
and puni shment s and to stop
pushing money into their faces.
My other concern is to empha-
size the futility of fiddling with
compensation schemes. Tbis is
not the road to quality. Andrew
Lebby, a consultant, and Eileen
Appelbaum, a researcher, corrob-
orate this, and each offers a way
of thinking about where excel-
lence actually comes from. I find
it useful to think in terms of three
C's: choice, collaboration, and
content. Choice means that em-
ployees should be able to partici-
pate in making decisions about
what tbey do every day. Collabo-
ration denotes tbe need to struc-
ture teams in order to facilitate an
exchange ot ideas and a elimate of
support. Content refers to wbat
people are asked to do: as Freder-
ick Herzberg said, "If you want
people motivated to do a good job,
give them a good job to do."
An organization that provides
these three ingredients in place of
artificial inducements like incen-
tive plans will not "lose its best
people," as Beer worries. Inno-
vation and excellence are the
natural results of helping people
experience intrinsic motivation.
But intrinsic motivation cannot
survive in an organization that
treats its employees like pets.
HARVARD BUSINESS REVIEW November-December 1993 49