Content uploaded by Benyamin B. Lichtenstein
Author content
All content in this area was uploaded by Benyamin B. Lichtenstein on Aug 02, 2016
Content may be subject to copyright.
Chapter 6—Positive Deviance Complexity and the Nexus of Leadership
Goldstein, Lichtenstein & Hazy
154
CHAPTER SIX
The Innovative Power Of Positive Deviance
In the early 1990s Vietnam was suffering from wide-scale child malnourishment, with
60-70% of the children in villages suffering from this life-threatening condition.i Since at that
time Vietnam was still not on very good relations with the United States—the war had been over
for less than twenty years—the Vietnamese government was not disposed to ask the US for help
with their social problems. Nevertheless, village leaders working with government officials did
turn to Save the Children; despite the charity’s connection to the US, the organization did have a
very good record in dealing with this particular malady. Save the Children, in turn, asked for
help from the late Jerry Sternin and his wife Monique Sternin, both of whom were long-time
NGO specialists and Jerry was fluent in Vietnamese. The Vietnamese government recognized
the gravity of the situation and aware of its own lack of funds to buy more food, put a lot of
pressure on Save the Children as well as the Sternins to show results in only six months.
From his many years working in underdeveloped and poor countries, Sternin knew that
the causes of malnourishment are systemic: ineffective food distribution systems; poor
sanitation; lack of potable water; low yielding agriculture; ignorance about new technologies that
could improve farming as well as food production; and others. He also knew the limitations of
the usual approach to alleviating malnourishment, the aid model, in which wealthy nations,
foundations or agencies provide money and other resources to purchase and distribute food.
though they can offer immediate relief, their successes are temporary —when the funding gives
out the problem remains behind and even gets worse since expectations have been raised only to
be smashed down. Moreover, since aid-based solutions are effectively ‘owned’ by the aid
workers rather than the people or communities who are suffering, these interventions de-energize
Chapter 6—Positive Deviance
155
the people being affected, reinforcing a passivity that reduces their capacity to work on
improvements to their own health behaviors. It is in such trying circumstances that the tools
coming out of complexity theory can be quite effective because the tried and true methods of the
past, the things that work in normative times and in equilibrium situations, are proving powerless
to effectuate constructive actions.
It was in this atmosphere that Sternin and his wife turned to a social intervention program
called “Positive Deviance” which they had adopted from research on childhood malnourishment
in developing countries, originally conducted by Marian Zeitlin at Tufts University.ii Zeitlin had
discovered something surprising: a small minority of the children were not malnourished, and
were doing much better than the rest of community. Using these cases as role models, she
labeled them positive deviants—“deviants” since they deviated from the vast majority of other
families, and “positive” since they were in fact doing better than the rest. Drawing on this idea,
the Sternins created a social intervention that sought out Positive Deviance conducted by
Positive Deviants. Specifically, working with village volunteers, they identified a minority of
community families which had deviated from the normal practices of village in terms of the way
they produced, gathered, cultivated, or ate their food, these deviations or, as we called them in
the previous chapter, “experiments in novelty, resulting in the positive outcome of raising
children whoe were not undernourished.
Overall, the results of this simple intervention were exhilarating. In Sternin’s words:
"It was wildly successful. We saw malnutrition drop 65% to 85% throughout the villages
in a two-year period. But that's not all that's thrilling: The Harvard School of Public
Health came to the four original villages and did an independent study. They found that
children who hadn't even been born when we left the villages were at the exact same
Chapter 6—Positive Deviance
156
enhanced nutritional levels as the ones who benefited from the program when we were
there. That means that the behavior sticks.”iii
The program eventually reached 2.2 million Vietnamese in 265 villages. Since then and over the
past decades, Save the Children and other NGOs have applied the positive deviance model to
help solve malnutrition in more than 20 countries, including Bangladesh, Bhutan, Bolivia,
Cambodia, Egypt, Ethiopia, Haiti, Myanmar, Nepal, and Sri Lanka. Further, Positive Deviance
has been applied in all sorts of social systems including neighborhoods, communities, small
firms, and even multinational corporations, where novel experiments are utilized to bring about
striking resolutions of long-standing, even what were thought to be intractable problems.iv
Positive Deviance provides a complexity-based, non-linear model for leading
constituencies through uncertain and changing environments by leveraging internal innovations
into system-wide change. This approach to leading innovation is decidedly different than
traditional views. It does not place the leader in the role of a visionary who looks out on the
situation confronting the organization, envisions a different future, and then effectively designs a
new kind of business model that positions the firm for that future. Moreover, Positive Deviance
does not inculcate laissez faire leadership in which “self-organization” is supposed to come
about if only enough hierarchical command and control infrastructures are dismantled.
Instead, the leadership model for Positive Deviance posits an active leadership role in
creating and facilitating the conditions for an ecology of innovation as we described in Chapter
2. In this chapter we extend the approach to experiments in novelty covered in the last chapter in
the direction of a practical leadership tool, that is, the specific nuts and bolts of how to institute a
Positive Deviance program to support the creation of an ecology of innovation. We also point out
the important differences between PD and other social interventions like benchmarking or best
Chapter 6—Positive Deviance
157
practices, which often don’t produce the results claimed for them precisely because of their lack
of a complexity science foundation.
Positive Deviance, Positive Deviants, And Experiments In Novelty
On the surface the themes this chapter—“Positive,” “Deviance,” and “Leading”—reflect
a paradox. “Deviance” has pejorative connotations since it points to behavior that is generally
considered socially noxious, yet it is juxtaposed with the term Positive, and linked to our view of
innovation. Moreover, what could “Deviance” have to do with “Leading,” except in the need to
eliminate deviance from the social nexus of a well-functioning organization?
Yet again, the science of complexity offers a unique view on social innovation which, as
we’ve said before, comes about neither through stronger top-down pressure nor stronger
pressures for group conformity. Instead, the generation of novelty is first initiated by deviances
from the mainstream functioning of an organization; then these deviances can go on to play a
very positive role but only if they are recognized, reinforced, and disseminated through
generative leadership.
In the previous chapter we presented how experiments in novelty, the very life blood of
an ecology of innovation, come out of the rich nexus of differences and micro-level diversity that
constitute a complex system—whether a large corporation, a small start up, an NGO, or a social
venture project like SEED. Research has shown that these experimental deviations are
ubiquitous, reflecting a power law distribution of fluctuations that characterizes dynamic,
complex systems. These pervasive experiments in novelty are considered to be “on the verge,”
because due to their minority status, their signals are typically too weak to be recognized as
seeds of innovation that they potentially are. However the very weakness of these signals makes
Chapter 6—Positive Deviance
158
them into a serious challenge for leaders not just to notice but, once noticed, how they can be
amplified into something useful for the purpose of innovation. It was in response to the difficult
challenge of weak signals, that we proposed the “intercohesion” social network structure, which
consists of the right mixture of strong tie and weak tie elements that are conducive for noting and
then acting on these deviant experiments in novelty.
Positive Deviance is a powerful intervention precisely because it incorporates all of these
features of experiments in novelty; moreover, it addresses the specific challenge of how to
identify, encourage and disseminate these deviations from the norm, especially in light of their
marginal status. This marginal status is what keeps deviant experiments in novelty from having
a larger impact, since there are strong pressures in every organization to dampen the voice (or
signals) of the minority. As pointed out by the organizational innovation researchers Ryan
Mathews and Watts Wacker, although many businesses may pay lip service to the need to
transgress the norm—in our language to establish an ecology of innovation—when it really
comes down to it, leaders “are terrified at the idea of deviating one degree away from the
formula that earned them past success.”v But who’s to blame them for their terror since,
according to conventional wisdom, if something worked in the past, then that seems to be a good
enough track record to justify doing more of it in the future.
In contrast, we have argued that innovation requires going beyond the norm; in fact, this
idea is embedded into the center of the word innovation—nova—which means doing new things
rather than repeating what worked in the past. From a complexity science perspective,
innovation often begins at the peripheries of the mainstream, through departures away from the
reigning conventional wisdom in an organization. And, the novelty generating potency of
Positive Deviance lies precisely in its marginal initial status. The goal, as we describe next, is to
Chapter 6—Positive Deviance
159
find a way to listen to these marginalized, weak signals, especially when these deviations from
the norm may be revolutionary in their implications. We turn then to how the core tenets of PD
can be applied by generative leadership in establishing ecologies of innovation in organizations.
The Program Of Positive Deviance
Assumptions and Objectives
At the core of a PD intervention is the recognition that significant innovation cannot
come through reliance on outside experts who, from their hierarchical command and control
position, tell the insiders what to do. Such a tactic in no way evokes the natural resources of the
system by leveraging the already effective practices of positive deviants within the system. In
contrast, the most fundamental tenet of PD, drawn from its focus on the natural resources of a
complex system, is that innovative solutions to long standing problems may be already present
within the community or organization. However, because those involved in generating these
solutions typically have only a marginal status, their solutions are not recognized by the rest of
the social system. Sternin encapsulated this core element of Positive Deviance in the following
principle:
! In every community there are certain individuals ("positive deviants") whose special
practices, strategies or behaviors enable them to find better solutions to prevalent community
problems than their peers who have access to the same resources.
This principle translates into the radical assumption that a small minority of individuals—
“positive deviants”—have the capacity to solve intractable problems or engage unprecedented
Chapter 6—Positive Deviance
160
opportunities in original and resourceful ways without recourse to special methods or external
resources. What makes these individual deviants is the manner in which their practices depart
from organizational norms and conventional wisdom; what makes them positive is that their
practices offer constructive solutions that the majority of members in the community have not
been able to solve. In other words, positive deviance represents experiments in novelty occurring
at the “fringe” of the mainstream but capable of being appropriated by that same mainstream.
Since the solution exists within small minority of members, leading through using PD
begins by gathering key players from the groups affected by the problem who then inquire into
their current practices as well as the problems or opportunities they are facing. In the Vietnam
example, villagers themselves would need to be involved from the start—individual families and
local leaders—and not just governmental officials or outside aid workers. In other words, what
Jerry and Monique had to do first was facilitate meetings between this diverse set of key players,
including villagers and their representatives, government workers, and aid experts. For example,
to ensure that there would be local ownership of the program, village health committees (VHCs)
were established, composed of members from the Women's and Farmer's Union, the People's
Committee, and other village health interested parties. The VHC then selected female volunteers
who were willing to make a commitment to the aims of the Positive Deviance program. Without
this sense of ownership all the positive outcomes could be attributed to “outside” experts, or in
corporate lingo, “top management” intervention.
In the case of Vietnam, the general problem was already known, viz., malnourishment of
children. But what about the details? What were the crucial elements of the problem? How did
they manifest in this system? The goal at this stage is to collect as many facts as possible about
the problem. In Vietnam this meant compiling health records that could show current data on
Chapter 6—Positive Deviance
161
degree of malnourishment, criteria for malnourishment, who exactly should be classified as
malnourished, etc. Again, the issue need not be a problem per se but could be an opportunity that
is not being recognized or taken advantage by the majority of community members.
Next comes in identifying the outliers, that is, the small minority of children and their
families who are not malnourished. The villagers did just that by carefully measuring the positive
deviants who were not malnourished and comparing that with careful measurements of the
majority, i.e., the children who were so. Rigorous measurement in this particular case would
include tabulating weights and heights and comparing these observations with health standards,
as well as measuring body fat using the same sort of calipers found in our over-nourished gym
nuts, plus other metrics of nourishment. The positive deviants would be defined as precisely
those children whose measurements indicated were not malnourished.
Note how different this move is when compared to the kind of traditional analysis that is
taught in business schools. As we described in the previous chapter, common statistics focus on
the averages of a population, using the Gaussian “normal distribution” as the model. Even our
language of looking for “outliers” reflects the negative view that most statistical analyses have of
cases that are not within the most common range of the distribution. In contrast, PD works in the
context of power law distributions—the complexity science measure of dynamic systems—
which highlights the very few cases that are extremely high in a certain dimension, in this case,
childhood nutrition. Those few cases can be distinguished from the 80%-90% of cases which are
malnourished. Rather than focus on the 80%, the real leverage point according to complexity
science, is the top 10%-20% of cases. This is precisely the focus of the PD intervention.
In Vietnam, this analysis sparked keen curiosity about what was responsible for these
children staying nourished, even though their families have the same access to food and other
Chapter 6—Positive Deviance
162
resources as the others in the village. This question reflects another key tenet of PD, namely that
things must be essentially equal among these positive deviants compared to their larger
communities. That is, if a family were receiving “care” packages of food from a rich uncle in Ho
Chi Minh City they would not be included in the program. A community member is only a
genuine positive deviant if they have they same access to resources as all the persons in the
social system that are touched by the problem or situation.
Once genuine positive deviants are identified, the focus shifts to uncovering what specific
practices are responsible for the lack of malnourishment of these positively deviant children.
This is accomplished the old fashioned way: by spending long periods of time with the positively
deviant children and families to discover exactly how their food and eating practices may differ
from the mainstream villagers. Here again, the intervention relies on villagers talking to villagers
rather than on ethnographic studies by outside experts on nutrition or food cultivation or in
statistical analysis by officials within the governmental hierarchy.
Over time, these committees gain sufficient information to understand what are the
practices which differ from the norm, and how these differ from other community members.
Next, attention shifts to how these novel practices can be disseminated—not just in the village
where the intervention is initiated but to all the villages who are facing a similar problem of
malnourishment. The end result must be the dissemination of practices which previously were
“deviant” but now become models of “positive” behaviors that can—and should—be taken up by
others in the social system, thereby replicating the approach and amplifying the deviance to solve
the problem for the greater community.
Thus, not only is the solution already present within the system, it is disseminated via the
same villagers involved in the project, i.e., the practices making-up this solution are spread from
Chapter 6—Positive Deviance
163
the minority to the majority of the system’s members by way of the villagers themselves. It is for
this reason that Jerry could emphasize how PD does not provoke an “immune reaction” or the
kind of resistance to change provoked by a “foreign” or external practice entering a system. In
Vietnam, this stage of PD had village families spending time with the positively deviant families
in each village, learning about the particular behaviors that make-up their successful resolution
of the problem of malnourishment.
In Vietnam, the specific practices identified as the key resolvers of the problem of
malnourished included the following experiments in novelty (note that the reasons why they
provided more nutrition were only accessible after the fact):
1. When cultivating rice off the patty (the primary staple of agriculture for these villagers is
rice cultivation), some of the sweet-potato greens around the rice were retained with the
rice as well. The greens provided extra minerals and vitamins as well as carbohydrates
that are critical for balanced nutrition.
2. Some of the small crustaceans (tiny shrimp and crabs) living in the water around the rice
plants were also included with the greens and the rice. These supplied extra protein and
calcium.
3. Note that the majority did not retain the greens nor the crustaceans since (a) the greens
were considered “low class” food, and (b) it was thought to be an inefficient way to
cultivate rice.
4. Another crucial element of positive deviant behaviors was that mothers tended to feed
their children more times per day. Accordingly, although the total amount of food for a
family was the same between the positively deviant families and the majority, the
practice of being fed more often during the day meant the children were on the whole
Chapter 6—Positive Deviance
164
eating more food relative to others since, if they were fed only once or twice a day they
simply couldn’t ingest that much at each time. As Jerry concluded from this finding,
“…using exactly the same amount of rice, spread out over an additional two or three
meals, the PD kids were getting more food and twice the calories as their neighbors who
had access to exactly the same resource!”
Eight Leadership Rules For Harnessing Positive Deviance
In an interview with Fast Company, Jerry Sternin emphasized eight crucial rules that
leaders can follow to ensure a successful PD program.vi These rules also make up a good list of
guidelines of how leadership can harness the power of experiments in novelty toward furthering
the creation of an ecology of innovation.
1. Don't presume that you have the answer. Of course, this presumption goes against nearly all
the conventional wisdom of leadership in which a sense of one’s own competence as a leader is
bound-up with believing you do indeed know the answer, or at least can get to someone in the
hierarchy or an outside expert who does. But in Positive Deviance, it critical to assume the very
opposite—indeed, your answer wouldn’t be accepted even if you did have it, because of your
very status as an outside “expert.” The key to positive deviance is that it has to emerge naturally
out of the community/organization.
2. Don't mix people from different communities, organizations, or departments. Again, we see a
major departure from the current leadership push for cross-functional teams that is heard so
loudly these days in corporate America. Of course, cross-functional teams can have great benefit
Chapter 6—Positive Deviance
165
in connecting across separate fiefdoms in corporations and other large organizations. But in PD,
according to Sternin, at the outset you are not at the point where generating a lively creative-
solving interchange would be of much help, although later on that could prove beneficial in
coming with ways of amping up the weak signal and distributing it through the community.
Instead, leadership needs to create a condition in which everyone is able to identify with the
others in the group because everyone is facing the same challenges and relying on the same set
of resources. This is also related to the critical point made above that everyone has to have equal
access to the same resources, i.e., all things are equal for all the parties concerned. Furthermore,
this like-mindedness of the groups’ participants will make it more probable that the positively
deviant practices will be accepted and then repeated by all social system members, and not just
by some differently motivated or skilled individuals with a different set of expertise.
3. Let the groups do it all themselves. Of course to get the thing rolling, the leaders need to
explain the situations and the tools, but then they must encourage and let the group itself decide
how they will go about gathering information on the problem/opportunity, and then let the group
identify the positive deviants and their effective practices. In addition, guidelines for how these
activities will be accomplished need to be established from the group members themselves. As
Sternin said, when the intervention in Vietnam began, they only went into four villages where
local women were trained to compile growth charts of age, weight, and height. It was these same
village women who were then asked if they knew of any children under three years of age who
came from the same poor families but were not malnourished. If they indeed know of such cases
of well nourished children they were asked, “You mean it's possible for a very poor family to
have a well-nourished child in this village?” The village women were not only amazed by their
Chapter 6—Positive Deviance
166
discovery, they were eager to immediately go find out what it was that these positively deviant
families were doing that was so different and more efficacious around nutrition than the
majority.
4. Identify conventional practices. Before positively deviant behaviors can be identified, you
have to first know what the current, accepted practices are. In the Vietnam case, Sternin
suggested that the volunteers observe how all the mothers in the village fed their children as well
as what attitudes the mothers had to eating that they were passing down to their children. The
village volunteers came up with the very interesting and important fact that certain foods were
considered too low-class or common although they were in fact nourishing. Also, they found
that, because of fear of diarrhea (a common cause of death by dehydration among children in
underdeveloped countries), some mothers tried to keep their children from eating very much at
all. Also, the villagers described how, because the parents were so busy, food was often just
lying around for children or they were fed only once or twice a day since it was too time
consuming. Remember, that as described above, in contrast the positive deviant children ate
small portions many times a day which enabled them to actually eat more food because “there's
only so much rice that a starving child's stomach can hold."
5. Identify and analyze the deviants. This stage is easier than may have been thought at first
impression since as the villagers tracked how all the community members went about their food
gathering, processing, and eating, a list of all the behaviors held in common as well as the
positive deviants and their positive deviance practices could naturally be seen. Moreover, at the
very same time, it became obvious that the small minority of positive deviants had to be doing
Chapter 6—Positive Deviance
167
something better since they had spontaneously solved the problem of malnourishment: their very
results proved it! Again, because it is the villagers themselves who will have found out all of this
by their own lights, they don’t need an outside solution to be imposed on them. It was the
villagers who accomplished all this and not the government officials nor the outside experts. As
Sternin stated, “They will have discovered a new way of doing things themselves, making it their
discovery, not yours. Analyze and list the set of behaviors that the deviants have in common.
Single out exactly what makes them successful.”
6. Let the deviants adopt deviations on their own. According to Sternin, rule six is critical for it is
quite different than the traditional approach to transferring knowledge. In particular,
disseminating positively deviant behavior is not about importing “best practices” from elsewhere
(we’ll say a lot more about what distinguishes PD from best practices later in the chapter), nor is
it a matter of reporting on new information. Instead, knowledge transfer in PD is about
disseminating new behaviors and new practices. As Sternin liked to say, it is new behavior, not
knowledge that is the source of innovation. Instead of teaching new knowledge, PD encourages
new behavior. “Let the people who have discovered the deviations spread the word in their
group. Don't require adherence to the new practices, but do offer incentives for it.” For example,
in Vietnam, a volunteer would invite eight mothers into her home, but the price of entry would
be that the mothers were required to bring a contribution of shrimp, crabs, and sweet-potato
greens. Then they would all cook with these ingredients along with the rice. This gathering
would continue for about two weeks; thereafter the mothers would do this in their own homes.
7. Track results and publicize them. Publicize the results in multiple ways including in these
Chapter 6—Positive Deviance
168
information broadcasts what the results are and how they were achieved, in order to let other
groups develop their own curiosity about them. Re-observe and post the new results on a
periodic basis and track the results quantitatively to demonstrate that the new results that come
with doing things differently.
8. Repeat steps one through seven. Here we can see evidence of the recursive operations so
crucial in complexity science. As Sternin pointed out, by repeating the whole thing in a cyclical
manner, “Chances are that they've discovered new deviations from the new norm. The bell curve
of performance keeps moving up, as long as you disseminate the best deviations across the curve
and continue to discover new examples of positive deviance among the next group of best
performers.”
Corporate Examples Of Positive Deviance In Action
Positive Deviance not only has been successful for dealing with long standing and what
had previously been thought to be intractable problems; PD appears to change the very culture of
the organizations or communities where it has been applied. As Jerry Sternin once heard from a
Bangladeshi village woman following on the heels of a PD program, "Let us tell you about the
changes in our lives. We were like seeds locked up in a dark place, and now we have found the
light."vii
These sentiments are not uncommonly spoken by participants in this kind of
social/organizational intervention. And lest it be thought that PD is only effective in dire
situations found in poor countries, in this section we will describe two examples of Positive
Deviance and its principles being applied in corporate settings. These examples concern two
Chapter 6—Positive Deviance
169
experiments in novelty at Merck Pharmaceuticals, certainly one of the premier companies in the
world in its industry. The first of these examples has to do with a spontaneous emergence of
novelty at the periphery, i.e., it is not about a formal program of PD being instituted at Merck.
However, the second example concerns an intentional PD intervention at Merck’s Latin
American headquarters in Mexico in which Jerry Sternin worked with chairman, and president of
the Plexus Institute in New Jersey, Henri Lipmanowicz and Curt Lindberg. The Plexus Institute
is dedicated to the application of the sciences of complex systems to improving health care
throughout the world, in terms of both organizational and health related issues (one of the
authors of this book, Goldstein, is a member of the science advisory board of the Plexus
Institute).Once we have gone over these examples, we will come back to more guidelines of how
generative leaders can use the principles of experiments in novelty in general and Positive
Deviance in particular.
Positive Deviance at Merck: Eradicating River Blindness
Merck, of course, is known for many medical breakthroughs as well as a recent bout
having to do with well-publicized problems with certain of its medicines. About thirty years ago,
the parasitologist Dr. William Campbell then working at Merck identified a substance that was
effective active against worms in livestock called ivermectin. The drug quickly became the most
successful animal medicine in history. Not only did it kill gastrointestinal worms in a single
dose, it also killed the biting insects that caused weeping sores on livestock. It had other uses as
well; if your dog takes monthly heartworm medication, its active ingredient is most likely
ivermectinviii.
When Campbell first described the invention to Dr. Mohammed Aziz—the primary
infectious disease specialist at Merck, they both quickly recognized the unique attributes of
Chapter 6—Positive Deviance
170
ivermectin. Aziz had worked in Africa for the World Health Organization and had seen many
cases of river blindness in humans, a very painful and disfiguring ailment found in various
regions in the developing world. The disease is caused when parasites growing under the human
skin escape, causing painful sores. When these parasites bore through the eyes, blindness
occurs—hence the term “river blindness.”
It turned out that ivermectin was effective on a related parasite in horses. Campbell and
Aziz approached Roy Vagelos, Merck’s future CEO who, at that time, as head of research at
Merck had established a reputation of openness and support for novel ideas and methods. We
have described other aspects of Vagelos’ complexity style of leadership in Chapter 2, particularly
his being a role model for creating communication channels among his social networks
according to what we have called the “interaction resonance”. The two scientists asked Roy for
further funding in order to develop and test a human version of ivermectin for treating river
blindness, and Vagelos agreed.
However, because river blindness is mostly confined to the developing world, the top
management at Merck realized they could not make up for their expenses by distributing the
drug for actual use. Also, there was the distinct possibility of side effects which might harm the
drug’s eventual reputation for use as a veterinary agent. Nevertheless, Vagelos and the other
senior leaders at Merck developed the drug at their own expense. In 1987 when the drug was
perfected, 18 million people around the globe were identified with river blindness.
But how can a corporation serve 18 million non-paying customers? After exhausting
governmental and global agency channels, Vagelos decided that Merck would act on its own,
significantly departing from the norms of the pharmaceutical industry. Merck chose to make and
distribute the drug for free to anyone at risk for river blindness anywhere around the world, and
Chapter 6—Positive Deviance
171
Vagelos pledged to do this for as long as it was necessary. The result significantly contributed to
the near eradication of this disabling disease, and continues to make important inroads against
related parasitic diseases such as Guinea worm. We see here positive deviance in full force:
positive outcomes emerging out of significant departures from the norm. And, of course, it didn’t
hurt Merck’s professional and public reputation!
PD at Merck’s Latin American Sales Headquarters
The second application of the idea of experiments in novelty at the periphery includes an
actual, intentional intervention of Sternin’s program of Positive Deviance for the Latin American
sales division again at Merck. The highlights of this story are told by Roberto Saco as part of his
master’s degree dissertation at Oxford University.ix Mr. Saco had access to Jerry Sternin, the
above mentioned representatives from the Plexus Institute, as well as the personnel at Merck
involved in the program including their leader Grey Warner.
Operating in Mexico since 1932, the Latin America group has a long-standing reputation
for both profitability and innovation. Indeed, Merck is known as one of the 20 best companies to
work for in Mexico. Under Grey Warner’s leadership, the Latin American Division had been
working at cultural transformation for a decade. One important ingredient of this objective is
called the “Living Dialogue” which aims at better understanding of the varied communities of
healthcare practice using Merck’s products and services. This has also included a commitment to
community engagement through establishing close collaborative relationships among the
members of these communities of practice, various NGOs whose mission dovetails with these
communities, as well as academic and governmental leadership in the geographical areas
affected.
Chapter 6—Positive Deviance
172
Warner Grey, who has been at the forefront of leading innovation at Merck,
avows,“…companies that want to succeed in today’s competitive environment need much more
[than product innovation]. They need innovation at every point of the compass, in all aspects of
the business and among every team member…. Fostering a culture of innovation is critical to
success.” Grey’s points about innovation, of course, mesh with the approach taken in this book
on establishing an ecology of innovation. Likewise, his goal to pursue innovation “at every point
of the compass” parallels the viewpoint that we have continuously expressed in the notion of
experiments in novelty that support an ecology of innovation. It is within Merck’s vision for
innovation that the PD program in the Sales Division was inaugurated in 2005 when Jerry
Sternin first visited Merck in Mexico.
The first product they examined was Fosamax, which fights against ostereoporosis, is one
of Merck’s giant products world-wide and in Mexico. A survey of the sales records in Mexico
revealed that only 22 out of the 180 sales representatives were gaining better-than-average sales.
Now, 22 divided by180 is only 12%; this small percentage should alert the leader to one of the
hallmarks of Positive Deviance, namely that some deviation (innovation) is likely to be at play
when a small minority is doing much better than the majority. These statistics, of course, prompt
the question of what were the minority doing differently than the rest. To be sure, in the
application of PD all other factors must be equal or it wouldn’t be a revealing statistic; for
example, the top 12% of sales reps might be focusing on wealthy neighborhoods with high
percentages of women getting older, or on unique government accounts. In this case, however,
the target population and other elements were exactly parallel.
So with these statistics in place, the next step was for a bottom-up identification on the
part of the sales representatives themselves of what was different in the sales practices of the
Chapter 6—Positive Deviance
173
positively deviant 12% of the Fosamax sales force. First, the usual practice (indeed, this seems
to be universal in the pharmaceutical industry) is to have sales representatives visit doctors’
offices on a regular basis. This face-to-face aspect is crucial to pharmaceutical sales, but it can
be irritating to doctors’ busy professional lives. However in Mexico, the sales representative
themselves actually do not close the sale -- instead, the product is actually bought and stocked
by a pharmacy. All the rep can do is influence the physician to prescribe the drug. The sales rep
is trained to “hit” each target or doctor several times per year with face-to-face selling but the
doctors find it time consuming and annoying.
This “design team” of sales employees did not come up with earth-shaking revelations
but they did find the following practices that were deviant, that is, were only practiced by the
minority “at the periphery”:
" Although many doctors only considered sales reps as mere vendors, the positively
deviant reps were instead perceived as “equals” by the doctors.
" Moreover, these same minority sales representatives had taken the time to build solid
partnering relationships with the docs by sharing information beyond simply product
specs, i.e., they were listening more than selling and selling their knowledge of the
product and not the product itself;
" they went deeper into pharmacy purchasing patterns including pharmacy prescription
audits to validate the doctors’ prescribing patterns;
" they localized their sales pitch;
" they made it a habit to call the physicians on their birthdays and other important events;
" they kept up to date on local current events to facilitate good conversations with the docs;
" the made use of company resources for arranging visits and other logistics;
Chapter 6—Positive Deviance
174
" they used satisfaction surveys at events;
" they checked pharmacy stocks pre- and post- events in order to gauge the impact of
presentations;
" they used visual aids to communicate the product’s advantages;
" they constantly repeated the product name as well as the word “new.”
By implementing many of these suggestions across the entire sales team, very strong
results were gained. Within a year Fosamax showed a 17% increase in sales; with a market
growing at 13% the entire division managed to reverse its trend of losing market share. Andres
Bruzual, the head of one of Merck’s three business units in Mexico, spoke highly of the PD
intervention: “A lot of learning is taking place across the board.” Furthermore, in a testament to
the power of the bottom-up process itself, Andres has warned that he would not initiate a
Positive Deviance program if the following conditions held: if managers felt they were the only
problem-solvers; if the groups were too small; if the circumstances were highly politicized; or if
internal competition was too high a value (which is certainly true in many sales situations).
Other overall learning coming out of the series of PD interventions are also relevant.
Participants realized that Positive Deviance should not be considered a one-shot program, but
rather must be part of a long-range and wide-spread initiative to foster an innovative internal
environment. This is precisely our point about systematizing novelty and adaptation by
establishing an ecology of innovation. Participants were also somewhat concerned that changes
were too sparse, an issue that Sternin himself complained about in terms of fragmented effort. In
that context, there was talk about the dangers of being too connected with only one product. As
a result they decided on “de-Fosomaxing” the process, i.e. encouraging novel practices with
other products and services.
Chapter 6—Positive Deviance
175
It is Easier to Practice than to Think Your Way to a New Behavior
One of Jerry Sternin’s favorite sayings about PD interventions is that it is easier to
practice rather than think your way to a new behavior. This was confirmed in Vietnam, at Merck,
and at the many other organizations and communities which have realized that novel practices at
the periphery can sometimes contain the seeds of innovation for an entire complex system.
However, those seeds take root only when they are planted—through new behaviors—rather
than simply talked about by the positive deviants. Note how both of these innovations at the
periphery—the health practices in Vietnam and the sales practices at Merck—are parallel to the
process a complex system undergoes in the emergence of new order. Fluctuations or
perturbations in a system can generate unexpected novelty that is a departure from the norm.
What begins as only an outlier effect can eventually become the seed that changes the whole
system. Whereas a management of control typically dampens or ignores such innovations,
generative leadership sees these deviations as a potential logic for a new order. This new logic,
if it can become embedded in the rest of the system, can make the entire organization it more
effective and adaptive.
How Positive Deviance Differs from Benchmarking
Although it might seem that by highlighting specific practices which work better than
others and then disseminating these practices throughout the system, Positive Deviance is just
another form of benchmarking. However, from both a complexity science perspective and Jerry
Sternin’s own insights, PD differs significantly in many respects to benchmarking. Certainly,
benchmarking has proven a very useful tool for improving organizational performance; our aim
is not to criticize it in general. Rather, by pointing to the important differences between PD and
Chapter 6—Positive Deviance
176
benchmarking we believe a surer grasp of the complexity aspects of Positive Deviance can be
had.
Simply put, benchmarking takes a work process or practice found effective in one
environment, and transplants it to another environment—in the belief that what worked in one
place will also work in the new one. However, as Sternin put it, this transplantation does not pay
enough heed to the contextual details of the place where the practice was initially developed. In
contrast, the PD approach focuses precisely on those very details within the same social system.x
The importance of context is a key finding within complexity science, with its concern on how
complex systems exhibit adaptability to their changing environments. What ‘context’ means is
precisely these variations in environments, which are always changing and thus confronting the
system with the need to stay abreast and adapt. Organizations are not only unique, possessing
different weaknesses and strengths so they shouldn’t be treated the same, their contextual
environments are also unique, changing according to their own patterns and rhythms.
Benchmarking simply fails to acknowledge this crucial fact about the relationship between
systems and their contexts.
Furthermore, whereas benchmarking tends to focus on identifying the best and most
effective process or approach with efficiency as the highest value, Positive Deviance does not
assume that efficiency is the most appropriate guide for adaptation to environmental change,
especially when efficiency is defined in terms of lower costs, higher production rates. Instead,
experiments of novelty at the periphery are all about new results, not ones measurable by some
pre-set standard of what the costs or time of production should be. Thus, whereas benchmarking
is about repeating what has previously proven successful, PD is about innovations that face
Chapter 6—Positive Deviance
177
toward to a new future. Such innovations are inherently idiosyncratic and hence are not
amenable, when they first appear, to analysis via efficiency considerations.
Also, whereas benchmarking brings ideas and concepts from outside to inside an
organization, Positive Deviance, as a method of identifying experiments in novelty at the
periphery assumes that innovation is already emerging inside the organization. Bringing
something from outside and imposing it on others will virtually always bring some measure of
resistance to change; therefore leaders using benchmarking must be ready with all sorts of tactics
to overcome this resistance to change. However, Positive Deviants are members of the same
group as the majority, and therefore their practices do not have to be imported. As we described
before, Sternin pointed out that the innovative practices discovered and disseminated by the
Positive Deviants have the advantage of not provoking an immune reaction from the host, since
they are part of the very host’s “body.” Experiments in novelty at the periphery are inherently
accessible to everyone in the social system because it is members of the same social system who
come up with them in the first place.
Sternin has pointed out that best practices face such a strong challenge of replication that
benchmarking often fails. Positive Deviance focuses on replicating the process, not the solution:
“[Solutions are] context-specific and contingent on resources, local conditions, politics,
and the like, and are therefore not universally adaptable. You cannot transplant a model
grown in one soil to another and expect it to flourish. Externally identified solutions
might require specific conditions that your business or organization does not offer.
Simply transferring the technical solution will not be sufficient.”
Chapter 6—Positive Deviance
178
While benchmarking is about replicating outcomes, Positive Deviance is about replicating the
process of learning new practices. The respect and ownership generated through this “process of
discovery” by those inside the problem and owning the solution is irreplaceable.
Conclusion: A Generative Leadership of Positive Deviance
In a conversation with Roberto Saco, Jerry Sternin clarified how he saw the role of
leadership in the utilization of Positive Deviance: the emphasis is on facilitation from the
perspective of a participant/observer and not on prescription.xi Indeed, for Positive Deviance to
work successfully, the leader/facilitator must in a sense blend into the background so that the
group members have an opportunity for learning within the community. In this capacity, Jerry
liked to quote from the Chinese sage Lao Tzu in his classic Tao de Ching:
Learn from the people
Plan with the people
Begin with what they have
Build on what they know.
Of the best leaders
When their task is accomplished
The people all remark
“We have done it ourselves.”
With this style in mind, let’s close the chapter by recapitulating the six “D’s” of Positive
Deviance which also apply to leading experiments in novelty in general as part of the strategy of
creating an ecology of innovation:
Define what the problem is, exactly, and determine the desired outcome. This relates to what we
have been calling Opportunity Tension. The leader identifies the natural stress on the
system that comes about when an opportunity is recognized in the environment by the
Chapter 6—Positive Deviance
179
members of an organization, but they also realize that, as currently configured, the system
is unable to realize its potential. Generative change is required.
Determine the individuals and groups of individuals who are already exhibiting the desired
outcome within the common behaviors in the community. These are the persons or
groups who represent the kind of novelty outside the norm that if replicated can change
the system. This confers on them the status of being Positive Deviants within the system.
Discover those unique practices or behaviors that enable the Positive Deviants to find better
solutions. This means carefully investigating exactly how and through what differences in
practices their novelty bestows advantages on those who practice them.
Design and implement social interventions whereby the majority have access to and learn the
new practices inherent in the novel successes of the Positive Deviants.
Discern the ongoing pulse of the organization by implementing mechanisms which provide
continuing monitoring and evaluation of the emergent novelty, initiatives and other
experiments across the organization.
Disseminate the successful innovations on as wide a scale as possible throughout the
organization. As changes to practice diffuse widely, system wide change, and emergent
order at a new level becomes possible as we described in Chapter Four.
Intriguingly, most of these elements reflect action across an entire social system—throughout the
social network. Understanding how social networks are structured can therefore support the
Chapter 6—Positive Deviance
180
process of PD, as well as further all the elements of an ecology of innovation. Thus, in the next
chapter, we explore the nature of social networks, and reveal key insights about the structure and
formation of networks for generative leadership.
Chapter 6—Positive Deviance
181
Notes:
i One of the authors of this book, Jeffrey Goldstein, had the good fortune of learning Positive Deviance
while working as a co-consultant with Jerry and Monique Sternin in 2006. Much of the material on PD in
this chapter is culled from the time Jeffrey spent with the Sternin’s on this project. The chapter is
ii Saco, R. (2005). Good companies: Organizations discovering the good in themselves by using Positive
Deviance as a change management strategy. MSc Dissertation, HEC Paris, Oxford Executive Education,
Oxford University.
iii Dorsey, D. (2000). Positive Deviant - Jerry Sternin. Fast Company, December. Available at the Plexus
Institute website: http://www.plexusinstitute.org/ideas/show_elibrary.cfm?id=270.
iv See the website devoted to Positive Deviance: http://www.positivedeviance.org.
v Mathews, R. & Wacker, W. (2002). The Deviant’s Advantage. NY: Random House., p. 240.
vi Dorsey, D. (2000). Positive Deviant - Jerry Sternin. Fast Company, December. Available at the Plexus
Institute website: http://www.plexusinstitute.org/ideas/show_elibrary.cfm?id=270.
vii Dorsey, D. (2000). Positive Deviant - Jerry Sternin. Fast Company, December. Available at the Plexus
Institute website: http://www.plexusinstitute.org/ideas/show_elibrary.cfm?id=270.
viii Vagelos, R. & Galambos, L. (2004). Medicine, Science, and Merck. Cambridge: Cambridge
University Press.
ix Saco, R. (2005). Good companies: Organizations discovering the good in themselves by using Positive
Deviance as a change management strategy. MSc Dissertation, HEC Paris, Oxford Executive Education,
Oxford University.
x Bertels, T. & Sternin, J. (2003). Replicating Results and Managing Knowledge. In Rath and Strong’s Six Sigma
Leadership Handbook. Hoboken, NJ: John Wiley and Sons, pp. 450-457.
Chapter 6—Positive Deviance
182
xi Saco, R. (2005). Good companies: Organizations discovering the good in themselves by using Positive
Deviance as a change management strategy. MSc. Dissertation, HEC Paris, Oxford Executive Education,
Oxford University, p. 19.
