Knowing When

to Pull the Plug

Barry M. Staw and Jerry Ross

Harvard Business Review

87212

This document is authorized for use only in Prof. Jatin Pandey,Prof. Kajari Mukherjee ,Prof. Sumit K. Ghosh &Prof. Srinath Jagannathan's PGP-I/Term-I/OB-I/AY: 2021-22 at Indian Institute of
Management - Indore from Jun 2021 to Sep 2021.
 HBR
MARCH-APRIL 1987

Knowing When to Pull the Plug

Barry M. Staw and Jerry Ross

Last year you authorized the expenditure of Of course, all managers will make some mistakes
$500,000 for what you thought was a promising new and stick with some decisions longer than they
project for the company. So far, the results have ought to. Recent research has shown, however, that
been disappointing. The people running the project the tendency to pursue a failing course of action is
say that with an additional $300,000 they can turn not a random thing. Indeed, at times some
things around. Without extra funding, they cry, managers, and even entire organizations, seem
there is little hope. Do you spend the extra money almost programmed to follow a dying cause.
and risk further losses, or do you cut off the project What leads executives to act so foolishly? Are they
and accept the half-million-dollar write-off? people who should never have been selected for respon-
Managers face such quandaries daily. They range sible positions? Are these organizations simply inept?
from developing and placing employees to choosing Or are they generally competent managers and compa-
plant sites and making important strategic moves. nies that find themselves drawn into decisional quick-
Additional investment could either remedy the situa- sand, with many forces driving them deeper? Though
tion or lead to greater loss. In many situations, a deci- we think this last description is probably the right one,
sion to persevere only escalates the risks, and good we don’t think the tendency is uncheckable. Managers
management consists of knowing when to pull the plug. and organizations that often fall into escalation traps
These escalation situations are trouble. Most of us can take steps to avoid them.
can think of times when we should have bailed out of
a course of action. The Lockheed L 1011 fiasco and the
Washington Public Supply System debacle (commonly
Why Projects Go Out of Control
referred to as WHOOPS) are spectacular examples of As a start to understanding why people get locked
organizational failure to do so. Decisions to persist into losing courses of action, let’s look first at what
with these crippled ventures caused enormous losses. a purely rational decision-making approach would

Mr. Staw is the Mitchell Professor of Leadership and

Communication at the Schools of Business Administration
at the University of California in Berkeley. He is also Mr. Ross is associate professor of organizational behavior at
chairman of its Organizational Behavior and Industrial the Graduate School of Industrial Administration at
Relations Group. Carnegie-Mellon University.

Copyright © 1987 by the President and Fellows of Harvard College. All rights reserved.

This document is authorized for use only in Prof. Jatin Pandey,Prof. Kajari Mukherjee ,Prof. Sumit K. Ghosh &Prof. Srinath Jagannathan's PGP-I/Term-I/OB-I/AY: 2021-22 at Indian Institute of
Management - Indore from Jun 2021 to Sep 2021.
 be. Consider, for example, the decision to pursue or benefits until the entire new system is completed.
scuttle an R&D or a marketing project. On the basis Unfortunately, as each year passes, the expected
of future prospects, you’d have made the initial deci- date of completion recedes into the future while the
sion to pursue the project, and enough time would bill for work to be finished grows exponentially. Of
have passed to see how things were going. Ideally, course, no one would have advocated the project if
you’d then reassess the situation and decide on the true costs had been known at the outset. Yet,
future action. If you were following a fully rational once begun, few have argued to kill the project.
approach, whatever losses might have occurred The problem is that the project was structured in
before this decision point would be irrelevant for ways that ensured commitment. First, the project
your reassessment. With a cold, clear eye, you’d managers viewed each setback as a temporary situa-
view the prospects for the future as well as your tion that was correctable over time with more
available options. Would the company be better off money. Second, they perceived all moneys spent as
if it got out, continued with the project, or decided investments toward a large payoff they’d reap when
to invest more resources in it? You’d treat any the project was complete. Third, expenditures were
previous expenses or losses as sunk costs, things irretrievable: the laid pipe in the ground has no
that had happened in the past, not to be considered value unless the entire project is completed, and it
when you viewed the future. would probably cost more to take the pipe out of the
In theory, pure rationality is great, but how many ground than it’s worth. Thus, like many other large
managers and organizations actually follow it? Not construction and R&D projects, investors in the
many. Instead, several factors encourage decision Deep Tunnel have been trapped in the course of
makers to become locked into losing courses of action. action. Even though what they receive in the end
may not measure up to the cost of attaining it, they
The Project Itself have to hang on until the end if they hope to recoup
any of their investment.
The first set of factors have to do with the project
itself. “Is the project not doing well because we Managers’ Motivations
omitted an important factor from our calculations,
or are we simply experiencing the downside of prob- Most of the factors concerning projects that
lems that we knew could occur?” “Are the problems discourage hanging on are evident to managers. They
temporary `bad weather or a soon-to-be-settled may not fully factor closing costs and salvage value
supplier strike’ or more permanent `a steep down- into their initial decisions to pursue certain courses of
turn in demand?’” Expected or short-term problems action (since new ventures are supposed to succeed
are likely to encourage you to continue a project. rather than fail), but when deciding whether to
You may even view them as necessary costs or continue a project or not, executives are usually aware
investments for achieving large, long-term gains. If of these factors. Less obvious to managers, however,
you expect problems to arise, when they do, they are the psychological factors that influence the way
may convince you that things are going as planned. information about courses of action are gathered, inter-
A project’s salvage value and closing costs can also preted, and acted on.
impede withdrawal. An executive could simply We are all familiar with the idea that people tend
terminate an ineffective advertising campaign in to repeat behavior if they are rewarded and to stop it
midstream, but stopping work on a half-completed if they are punished. According to the theory of
facility is another story. A project that has very little reinforcement, managers will withdraw from a
salvage value and high closing costs—payments to course of action in the face of bad news. This inter-
terminated employees, penalties for breached pretation, however, ignores people’s history of
contracts, and losses from the closing of facilities— rewards. Managers have often been rewarded for
will be much more difficult to abandon than a project ignoring short-run disaster, for sticking it out
in which expenditures are recoverable and exit is easy. through tough times. Successful executives—
It’s understandable why so many financially question- people whose decisions have turned out to be
able construction projects are pursued beyond what winners even when the outlook had appeared grim—
seems to be a rational point of withdrawal. are particularly susceptible. It’s tough for managers
Consider the Deep Tunnel project in Chicago, a with good track records to recognize that a certain
plan to make a major addition to the city’s sewer course isn’t a satisfactory risk, that things aren’t
system that will eventually improve its capacity to once again going to turn their way.
handle major storms. Although the project has Reinforcement theory also tells us that when
absorbed millions of dollars, it won’t deliver any people receive rewards intermittently (as from slot

HARVARD BUSINESS REVIEW March-April 1987 2

This document is authorized for use only in Prof. Jatin Pandey,Prof. Kajari Mukherjee ,Prof. Sumit K. Ghosh &Prof. Srinath Jagannathan's PGP-I/Term-I/OB-I/AY: 2021-22 at Indian Institute of
Management - Indore from Jun 2021 to Sep 2021.
 machines), they can become quite persistent. If a however, involve some additional factors that come
decline in rewards has been slow and irregular, a into play when other people are around to observe
person can go on and on even after the rewards have our actions. These are social determinants.
disappeared. Unfortunately, many business situa-
tions that escalate to disaster involve precisely this Social Pressures
type of reinforcement pattern. Sales may fall slowly
in fits and starts, all the while offering enough hope Managers may persist in a project not only
that things will eventually return to normal. The because they don’t want to admit error to them-
hope makes it difficult to see that the market may selves but also because they don’t wish to expose
have changed in fundamental ways. Revenues that their mistakes to others. No one wants to appear
slowly sour or costs that creep upward are just the incompetent. Though persistence may be irrational
kind of pattern that can cause managers to hang on from the organization’s point of view, from the point
beyond an economically rational point. of view of the beleaguered manager seeking to
Research has also shown other reasons that exec- justify past behavior, it can be quite understandable.
utives fail to recognize when a project is beyond When a person’s fate is tied to demands for perfor-
hope. People have an almost uncanny ability to see mance and when accepting failure means loss of
only what accords with their beliefs. Much like power or loss of a job, hanging on in the face of
sports fans who concentrate on their own team’s losses makes sense. Research has shown, for
great plays and the other team’s fouls, managers example, that job insecurity and lack of managerial
tend to see only what confirms their preferences. support only heighten the need for external justifi-
For example, an executive who is convinced that a cation. Thus when a manager becomes closely iden-
project will be profitable will probably slant esti- tified with a project (“that’s Jim’s baby”), he can be
mates of sales and costs to support the view. If the essentially forced to defend the venture despite
facts challenge this opinion, the manager may work mounting losses and doubts about its feasibility
hard to find reasons to discredit the source of infor- Beyond the personal risks of accepting losses, our
mation or the quality of the data. And if the data are ideas of how a leader should act can also foster foolish
ambiguous, the manager may seize on just those persistence. Culturally, we associate persistence—
facts that support the opinion. Thus information ”staying the course,” “sticking to your guns,” and
biasing can be a major roadblock to sensible with- “weathering the storm”—with strong leadership.
drawal from losing courses of action. Persistence that happens to turn out successfully is espe-
In addition to the effects of rewards and biased cially rewarded. For example, when we think about the
information, a third psychological mechanism may be people who have become heroes in business and politics
at work. Sometimes even when managers recognize (Iacocca and Churchill, for examples), we see leaders
that they have suffered losses, they may choose to who have faced difficult and apparently failing situa-
invest further resources in a project rather than accept tions but who have hung tough until they were
failure. What may be fostering escalation in these successful. If people see persistence as a sign of leader-
cases is a need for self-justification. Managers may ship and withdrawal as a sign of weakness, why would
interpret bad news about a project as a personal they expect managers to back off from losing courses of
failure. And, like most of us who are protective of our action? Recent research demonstrates that even though
self-esteem, managers may hang on or even invest it may not add to the welfare of the organization, persis-
further resources to “prove” the project a success. tence does make a manager look like a leader.
A number of experiments have verified this effect of In short, the need to justify one’s actions to others
self-justification. Those who are responsible for and to appear strong as a leader can combine with
previous losses, for example, have generally been found the three psychological factors to push managers
to view projects more positively and to be more likely into staying with a decision too long. This combi-
to commit additional resources to them than are nation of forces does not, however, account for all
people who have taken over projects in midstream. debacles in which organizations suffer enormous
Managers who are not responsible for previous losses losses through excessive commitment. In many of
are less likely to “throw good money after bad” since these cases structural factors also play a role.
they have less reason to justify previous mistakes.
Reinforcement, information biasing, and self- Organizational Pushes & Pulls
justification—three psychological factors that we’re
all subject to—can keep us committed to projects or Probably the simplest element impeding with-
actions we have started. Most managerial decisions, drawal from losing projects is administrative inertia.

3 HARVARD BUSINESS REVIEW March-April 1987

This document is authorized for use only in Prof. Jatin Pandey,Prof. Kajari Mukherjee ,Prof. Sumit K. Ghosh &Prof. Srinath Jagannathan's PGP-I/Term-I/OB-I/AY: 2021-22 at Indian Institute of
Management - Indore from Jun 2021 to Sep 2021.
 Just as individuals do not always act on their beliefs, The problem was not ending the project per se but
organizations do not always base their practices on what it symbolized. The L 1011 was Lockheed’s
their preferences. All the rules, procedures, and major entry in the commercial aviation market (in
routines of an organization as well as the sheer which it had been a pioneer), and Lockheed shrank
trouble it takes for managers to give up day-to-day from being identified as simply a defense contractor.
activities in favor of a serious operational disruption Pan American World Airways has recently gone
can cause administrative inertia. Dropping a line of through a similar institutional process. More than
business may mean changing corporate layoff poli- most airlines, Pan Am suffered huge losses after
cies, and moving people to other projects may deregulation of the industry; it was even in danger of
violate seniority and hiring procedures. Sometimes not meeting its debt obligations. Although the
it’s just easier not to rock the boat. prospects for large profits in the airline industry
Beyond such simple inertia, the politics of a situ- were dim, Pan Am chose to sell off most of its other
ation can prevent a bailout. British Columbia’s deci- more profitable assets—first the Pan Am building in
sion to stage the world’s fair Expo ‘86 is one of the New York and then the Intercontinental Hotels
most recent public examples of the power of polit- Corporation—so as to remain in its core business.
ical forces to sustain a costly course of action. Expo Finally, as losses continued, Pan Am sold its valu-
‘86 was supposed to operate close to the financial able Pacific routes to United Air Lines. Following
break-even point. But as plans for the fair got under these divestitures, the company was left with only
way, the expected losses burgeoned. At first, the U.S. and international routes in corridors where
planners tried to minimize the financial hazards by competition is heavy. Apparently, management
providing heartening but biased estimates of didn’t seriously consider the possibility of selling or
revenues and costs. When they finally accepted the closing the airline and keeping most of the other
more dire financial projections, however, and even profitable subsidiaries. Pan Am is, after all, in the
the director recommended cancellation, the plan- airline and not the real estate or hotel business.
ners still went ahead with the fair. Politically it was Not all the forces we’ve described are relevant to
too late: the fortunes of too many businesses in the every case, and not all are of equal influence in the
province were tied to Expo, it was popular with the situations where they operate. In many instances,
voters, and the future of the premier and his polit- commitment to a course of action builds slowly.
ical party were aligned with it. The province created Psychological and social forces come into play first,
a lottery to cope with the expected $300 million and only later does the structure make its impact.
deficit, and the fair opened as scheduled. And, in a few cases, because the rational point of
Though the Expo example comes from the public withdrawal has long passed, even the economic
sector, political force may also sustain costly business aspects of a project can cry out for continuation.
projects. As a venture withers, not only those Still, some executives do manage to get them-
directly involved with it may work to maintain it, but selves and entire organizations out of escalating
other interdependent or politically aligned units may situations. There are solutions.
support it as well. If the project’s advocates sit on
governing bodies or budget committees, efforts to
stop it will meet further resistance. If a review finally
Steps Executives Can Take Themselves
does occur, the estimates of the costs and benefits of Executives can do many things to prevent
continuing the venture will very likely be biased. becoming overcommitted to a course of action.
On occasion, support for a project can go even Some of these solutions they can take care of on
deeper than administrative inertia and politics. their own. Others involve getting the organization
When a project such as a long-standing line of busi- to do things differently. Let’s look first at the reme-
ness is closely identified with a company, to dies that executives themselves can apply.
consider its discontinuation is to consider killing
the very purpose of the company. (Imagine Hershey Recognize Overcommitment
without chocolate bars or Kimberly-Clark without
Kleenex.) A project or a division can become institu- The most important thing for managers to realize
tionalized in an organization. is that they may be biased toward escalation. For all
Consider the plight of Lockheed with its L 1011 the reasons we have mentioned, executives may
Tri-Star Jet program. Although every outside delude themselves into thinking that a project will
analysis of the program found the venture unlikely pull through—that success is around the corner.
to earn a profit, Lockheed persisted with it for more Recognizing overcommitment is, however, easier to
than a decade and accumulated enormous losses. preach than to practice. It usually takes enthu-

HARVARD BUSINESS REVIEW March-April 1987 4

This document is authorized for use only in Prof. Jatin Pandey,Prof. Kajari Mukherjee ,Prof. Sumit K. Ghosh &Prof. Srinath Jagannathan's PGP-I/Term-I/OB-I/AY: 2021-22 at Indian Institute of
Management - Indore from Jun 2021 to Sep 2021.
 siasm, effort, and even passion to get projects off the “nothing but the truth” policy, however, is usually
ground and running in bureaucratic organizations. not enough to change the pattern of information
The organization depends on these responses for reporting. The messenger with extremely critical
vitality. Consequently, the line between an opti- but important information needs an explicit reward.
mistic, can-do attitude and overcommitment is very One forum for getting objective and candid feed-
thin and often difficult to distinguish. back is a variant of the currently popular quality
circle. Managers could regularly convene key staff
See escalation for what it is. How, then, can managers members for “decision circles,” in which fellow
know whether they have crossed the threshold employees would offer honest evaluations of the
between the determination to get things done and hurdles a project faces and its prospects. Managers
overcommitment? Although the distinction is often from other departments or sections might also
subtle, they can clarify matters by asking them- attend or even chair such sessions to ensure an
selves the following questions. objective look at the problems. Managers might also
hold regular “exchanges of perspective” in which
1. Do I have trouble defining what would constitute colleagues could help each other see the truth about
failure for this project or decision? Is my definition their operations.
of failure ambiguous, or does it shift as the project
evolves? Change the Organization
2. Would failure on this project radically change the
way I think of myself as a manager or as a person? Though it is possible to come up with an array of
Have I bet the ranch on this venture for my career or decision aids to help managers gain an objective
for my own satisfaction? perspective about the projects they run, one could
3. Do I have trouble hearing other people’s concerns argue that the problem of escalation is larger than
about the project, and do I sometimes evaluate any one person, that it’s organizational in scope.
others’ competence on the basis of their support for Unfortunately, such a pessimistic view is at least
the project? partially correct. Much of what causes escalation is
4. Do I generally evaluate how various events and in the nature of organizations, not people.
actions will affect the project before I think about If organizational action is called for, what can the
how they’ll affect other areas of the organization or system do to minimize escalation?
the company as a whole?
5. Do I sometimes feel that if this project ends, Turn over administrators. One way to reduce the
there will be no tomorrow? commitment to a losing course of action is to
replace those associated with the original policy or
If a manager has answered yes to one or more of project. If overcommitment stems from psycholog-
these questions, the person is probably overcom- ical and social forces facing the originators of the
mitted to a project. action, then their removal eliminates some of the
sources of commitment.
Back off. Just knowing that one is under the sway of Turning over project managers can of course be
escalation can help. But knowing is not enough. It both disruptive and costly. Moreover, because
is also necessary to take some steps to avoid over- people who were once associated with the discon-
commitment. One way is to schedule regular times tinued venture may still be committed to it,
to step back and look at a project from an outsider’s management may find it difficult to draw the appro-
perspective. A good question to ask oneself at these priate line for making a purge. Nonetheless, to
times is, “If I took over this job for the first time make a clean break with the past, many organiza-
today and found this project going on, would I tions do make occasional personnel sweeps, some-
support it or get rid of it?” Managers could take their times more for their symbolic value than because of
cues from bankers. When they take over others’ any real differences in decision making.
portfolios, bankers usually try to clean up any trou- Still, we don’t recommend turnover as the way to
bled loans since they want to maximize the future make changes. Like treating the disease by killing
returns associated with their own loan activity. the patient, taking committed decision makers off a
Managers can also encourage their subordinates to project may produce nothing but a demoralized staff
reevaluate decisions. Most critical here is estab- and disaffected managers hesitant to try again.
lishing a climate in which, regardless of whether the
data are supportive or critical of the ongoing project, Separate decision makers. One technique for reducing
people convey accurate information. Just stating a commitment that is far less drastic than turnover is to

5 HARVARD BUSINESS REVIEW March-April 1987

This document is authorized for use only in Prof. Jatin Pandey,Prof. Kajari Mukherjee ,Prof. Sumit K. Ghosh &Prof. Srinath Jagannathan's PGP-I/Term-I/OB-I/AY: 2021-22 at Indian Institute of
Management - Indore from Jun 2021 to Sep 2021.
 separate initial from subsequent decisions concerning Because no one wants to be the conveyer of bad
a course of action. In some banks, for example, a news, information is filtered as it goes up the hier-
“workout group” handles problem loans rather than archy. Furthermore, because those intimately
the people who originally funded and serviced the involved with a project are not likely to distribute
loans. The idea is not only that specialists should be unflattering and less-than-optimistic forecasts,
involved in recouping bank funds but also that these information is also biased at the source.
officers are able to handle the loans in a more objective What, then, can organizations do to improve their
way than those who made the first decisions about the information reporting? The most common solution
accounts. Industrial companies could also make use of is to increase their use of outside experts and consul-
such procedures. They could separate funding from tants. The problem with consultants, however, is
new-product-development decisions and hiring from that they are no more likely to hear the truth than
promotion decisions. They could keep deliberations anyone else in the organization, and they also may
on whether to discontinue lines of business apart from not find it easy to tell management what it doesn’t
day-to-day management decisions. want to hear.
A better solution is to try to improve the honesty of
Reduce the risk of failure. Another way to reduce reporting throughout the organization. By rewarding
commitment is to lessen the risk of failure. Because process as highly as product, managers can encourage
project failure can spell the end to an otherwise candid reporting. The purpose of rewarding managers
promising career, an administrator may be forced to for the way a process is carried out is to make them
defend a losing course of action. In a no-win dilemma, attend as much to the quality of analysis and decision
the trapped manager may think, “Things look bleak making as to the final results. Instead of acting as
now, but there’s no point in my suggesting that the champions who inflate the prospects of their own
company withdraw. If the project doesn’t succeed, I projects and minimize their risks, managers offered
have no future here anyway.” process rewards are motivated to recognize problems
In some companies, management has reduced the and deal with them..
costs of failure by providing rationalizations for At the outset of projects, companies should
losing courses of action and excuses for their encourage the creation of fail-safe options, ways to
managers. People are told that the losses are beyond segment projects into small, achievable parts, and
anyone’s control or that the fault lies with more analyses of the costs of withdrawal. Later in the life
general economic conditions, government regula- of projects, companies should reward honest recog-
tion, or foreign competition. Although this route nition of problems and clear examination of the
takes managers off the hook, it doesn’t help them alternatives, including withdrawal.
see a losing course for what it is or how they may This kind of reward system is quite different from
avoid making the mistakes again. the usual practice of giving people recognition for
Most companies do not want to take the pressure success on their projects and punishing them for
off their managers to perform as winners. Yet failure on their undertakings. Yet it is a system that
because a strong fear of failure can cause overcom- should reduce many of the forces for escalation.
mitment, management is better off setting only a
moderate cost for failure, something to avoid but not Boosting Experimentation
to fear intensely. A large computer company, for
example, puts managers who have made big As we noted earlier in our discussion, an entire
mistakes in a “penalty box.” It makes them ineli- organization can be so caught up in supporting a
gible for major assignments for up to a year. After project—especially an institutionalized one—that it
the penalty period, the managers are restored to full ignores the cost of persistence.
status in the organization and are again eligible to Rather than trying to discredit an institutionalized
run major projects. Organizations trying to cope project on economic grounds, a good strategy for
with escalation situations may find such a compro- withdrawal from it is to reduce its links with the
mise between support for failure and demand for central purposes of the organization. A useful tactic is
competence helpful. to label the project peripheral or experimental so that
managers can treat it on its own merits rather than as
Improve the information system. Several laboratory a symbol of the organization’s central goal or mission.
experiments have shown that people will withdraw Ideally, managers should consider all ventures
from escalating situations when they see the high imperfect and subject to question in an “experi-
costs of persisting. The presentation of such nega- menting organization.” Every program should be
tive data is more difficult in organizations, however. subject to regular reconsideration (a la zero-based

HARVARD BUSINESS REVIEW March-April 1987 6

This document is authorized for use only in Prof. Jatin Pandey,Prof. Kajari Mukherjee ,Prof. Sumit K. Ghosh &Prof. Srinath Jagannathan's PGP-I/Term-I/OB-I/AY: 2021-22 at Indian Institute of
Management - Indore from Jun 2021 to Sep 2021.
 budgeting), and every line of business should be for units and how they cope with them as for success
sale at the right price. In such an experimenting and failure, experimenting organizations should be
organization, projects wouldn’t become institution- extremely flexible. When a market or a technology
alized to the point where management couldn’t changes, the experimenting organization would not
judge them on their own costs and benefits. And simply try to patch up the old product or plant but
because managers in such a system would be judged would be quick to see when it is best to pull the plug
as much for recognition of problems facing their and start anew.

References

1 For more complete reviews of escalation research, see Barry M. Staw and Jerry Ross, “Understanding Escalation Situations: Antecedents, Prototypes,
and Solutions,” in Research in Organizational Behavior, ed. L.L. Cummings and Barry M. Staw (Greenwich, Conn.: JAI Press, 1987); and Joel Brockner
and Jeffrey Z. Rubin, Entrapment in Escalating Conflicts (New York: Springer-Verlag, 1985).

2 See Gregory B. Northcraft and Gerrit Wolf, “Dollars, Sense, and Sunk Costs: A Lifecycle Model of Resource Allocation,” Academy of Management
Review, April 1984, p. 22.

3 For experiment results, see Barry M. Staw, “Knee-deep in the Big Muddy: A Study of Escalating Commitment to a Chosen Course of Action,” in
Organizational Behavior and Human Performance, June 1976, p. 27; Alan Tegar, Too Much Invested to Quit (New York: Pergamon Press, 1980); Max
H. Bazerman, R.I. Beekum, and F. David Schoorman, “Performance Evalution in a Dynamic Context: A Laboratory Study of the Impact of Prior
Commitment to the Ratee,” Journal of Applied Psychology, December 1982, p. 873.

4 Frederick V. Fox and Barry M. Staw, “The Trapped Administrator: The Effects of Job Insecurity and Policy Resistance
upon Commitment to a Course of Action,” Administrative Science Quarterly, September 1979, p. 449.

5 Barry M. Staw and Jerry Ross, “Commitment in an Experimenting Society: An Experiment on the Attribution of Leadership from Administrative
Scenarios,” Journal of Applied Psychology, June 1980, p. 249.

6 See Roy J. Lewicki, “Bad Loan Psychology: Entrapment and Commitment in Financial Lending,” Graduate School of Business Administration Working
Paper No. 80-25 (Durham, N.C.: Duke University, 1980).

7 Bruce E. McCain, “Continuing Investment Under Conditions of Failure: A Laboratory Study of the Limits to Escalation,” Journal of Applied
Psychology, May 1986, p. 280; and Edward G. Conlon and Gerrit Wolf, “The Moderating Effects of Strategy, Visibility, and Involvement on Allocation
Behavior: An Extension of Staw’s Escalation Paradigm,” Organizational Behavior and Human Performance, October 1980, p. 172.

8 Donald T. Campbell, “Reforms as Experiments,” American Psychologist, April 1969, p. 409.

7 HARVARD BUSINESS REVIEW March-April 1987

This document is authorized for use only in Prof. Jatin Pandey,Prof. Kajari Mukherjee ,Prof. Sumit K. Ghosh &Prof. Srinath Jagannathan's PGP-I/Term-I/OB-I/AY: 2021-22 at Indian Institute of
Management - Indore from Jun 2021 to Sep 2021.
 Harvard Business Review
HBR Subscriptions Harvard Business Review
U.S. and Canada Outside U.S. and Canada
Subscription Service Tower House
P.O. Box 52623 Sovereign Park
Boulder, CO 80322-2623 Lathkill Street
Market Harborough
Leicestershier LE16 9EF
Telephone: (800) 274-3214 Telephone: 44-85-846-8888
Fax: (617) 496-8145 Fax: 44-85-843-4958
American Express, MasterCard, Visa accepted. Billing available

HBR Article Reprints Harvard Business School Publishing

HBR Index and Other Catalogs Customer Service - Box 230-5
HBS Cases 60 Harvard Way
HBS Press Books Boston, MA 02163
Telephone: U.S. and Canada (800) 545-7685
Outside U.S. and Canada: (617) 495-6117 or 495-6192
Fax: (617) 495-6985
Internet address: custserv@cchbspub.harvard.edu

HBS Management Productions Harvard Business School Management Productions videos are
Videos produced by award winning documentary filmmakers. You’ll find
them lively, engaging, and informative.

HBR Custom Reprints Please inquire about HBR’s custom service for quantity discounts.
We will print your company’s logo on the cover of reprints, collections,
or books in black and white, two color, or four color. The process is
easy, cost effective, and quick.

Telephone: (617) 495-6198 or Fax: (617) 496-8866

Permissions For permission to quote or reprint on a one-time basis:

Telephone: (800) 545-7685 or Fax: (617) 495-6985

For permission to re-publish please write or call:

Permissions Editor
Harvard Business School Publishing
Box 230-5
60 Harvard Way
Boston, MA 02163
Telephone: (617) 495-6849

This document is authorized for use only in Prof. Jatin Pandey,Prof. Kajari Mukherjee ,Prof. Sumit K. Ghosh &Prof. Srinath Jagannathan's PGP-I/Term-I/OB-I/AY: 2021-22 at Indian Institute of
Management - Indore from Jun 2021 to Sep 2021.