You are on page 1of 9

Technology in Society 33 (2011) 244252

Contents lists available at SciVerse ScienceDirect

Technology in Society
journal homepage: www.elsevier.com/locate/techsoc

The 2010 BP Gulf of Mexico oil spill: Implications for theory


of organizational disaster
Barry Bozeman
Department of Public Administration and Policy, University of Georgia, 201 Baldwin Hall, Jackson, St. Athens, GA 30602, United States

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 10 August 2011
Received in revised form 28 September 2011
Accepted 28 September 2011

Focusing on the interaction of technology and organizational factors, the present paper
examines the 2010 BP Gulf of Mexico oil spill for the purpose of developing a better
understanding of the requirements for a theory of organizational disasters. Drawing from
literature on organizational disasters, a model of technology-embedded disasters is
developed and discussed. After outlining the events surrounding the oil spill disaster, the
model is employed in analysis of the oil spill. The oil spill case is employed as a means of
reecting on the requirements for an improved model of organizational disaster.
2011 Elsevier Ltd. All rights reserved.

Keywords:
Organizational disaster
Technology
Oil spill
Theory of catastrophic events

1. Introduction
One of the worst environmental disasters in world history
began in the early evening of April 20, 2010 as escaping,
uncontrolled hydrocarbons caused an explosion on the oil
drilling platform Deepwater Horizon, a facility leased by BP
Exploration & Production Incorporated from Transocean, an
international technology company providing rig-based, wellconstruction products and services. Within 36 h the gigantic
oil rig, approximately 300 by 300 feet and large enough to
provide living quarters for about 175 people, complete with
cafeteria, recreation facilities and helicopter pad, had crumbled and sunk into the ocean. Eleven inhabitants on the rig
died, most as a result of the explosion, and 17 others were
injured. The damage had only just begun. Crude oil gushed
from the uncapped well until July 15, 2010 and it was not until
September 19, 2010 that the well was completely sealed [52].
As the oil spill spread out to cover more than
88,000 square miles and made its way to beaches and
estuaries, incalculable damage was done to tourism and
shing industries and to the many marine animal and bird
species in and around the Gulf of Mexico [47]. By BPs

E-mail address: bbozeman@uga.edu.


0160-791X/$ see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.techsoc.2011.09.006

estimate [26], the oil spill cost the company more than $40
billion in both direct outlays, including a $20 billion victim
compensation fund.
During the six months following the BP oil spill media
coverage was so ubiquitous that perhaps historian Douglas
Brinkley was correct in his assessment that the general
public had experienced an oil spill fatigue [9]. While not
seeking to resuscitate oil spill fatigue, I provide yet another
analysis, one focused on the interaction of technology and
organizational factors and inter-organizational relations. I
provide no holistic analysis of the disaster in all its aspects
but, instead, use the disaster as a building block to further
understanding of the interactions of organizations and
technology in disasters [12]. The model presented here,
termed the technology-embedded disaster model, will
perhaps shed some small light on those disasters where
organizations and technology interact, but the model will
not account fully for diverse causes of the BP oil spill. It is
not my purpose to identify causal factors and then parse
them out according to their degree of culpability. Sufce to
say that the causes were many, diverse and interrelated and
that others (see especially [52] have already provided
causal analyses that are as exacting as we are likely, at least
in the near term, to obtain. My focus on the BP oil spill is
more as a case in point, using the case as a backdrop to

B. Bozeman / Technology in Society 33 (2011) 244252

considering the need and possibilities for a better theoretical understanding of organizational disasters.
Inasmuch as the term technology plays a pivotal role
in this analysis, it is perhaps useful to begin by dening it.
My denition is a physical embodiment of knowledge
designed to in its application solve a problem or serve
a perceived need. In this usage, technology cannot be
only social; some physical embodiment is required. A
further elaboration of this denition can be found in [14].
The next section of the paper considers organizational
disaster as a topic for social science theory. After this
context setting, I discuss a technology-embedded disaster
model, a model designed to be applicable to a broad array
of organization disasters. Then the paper turns to the BP oil
disaster and its particulars, focusing specically on deep
sea drilling technologies and the events, both the idiosyncratic events and the institutional and organizational
contexts of those events. Finally, I refer to the technologyembedded disaster model and seek to determine the
ways in which the BP oil disaster has broader implications
for the interaction of technologies and organizational
cultures in disaster creation.
2. Organization disaster as a theory-building
challenge
When cataclysmic events occur, vast reportage and
government investigations follow soon afterward. This has
certainly been the case with the BP oil spill, as evidenced by
an outpouring of books (e.g. [22,27,34,64], media reports,
government reports [52] and scholarly articles [1,49]; Pade,
2010). Early analyses of the BP oil spill have varied in causal
attribution focusing variously on corporate greed, poor
government regulation, idiosyncratic factors, or on the
worlds addiction to oil.
Other disasters have generated ne works on the
interaction of technology and organizations, many focusing
on decision-making (e.g. [3,37]. From these works, we

245

know more about how bad or unfortunate decisions led to


catastrophic outcomes in such cases as the Challenger
disaster [63], Three Mile Island, the Bhopal chemical spill
[10], and the Exxon-Valdez oil spill [66].
While there is no shortage of formal knowledge about
organizational disasters, there is a large body of literature
on organizational disasters that is multidisciplinary and
diverse in its assumptions, approaches and methods.
Literature on organizational disaster is characterized by
a variety of approaches and foci including, among others,
leadership failures [38,41] and failures due to external
contingencies [17,29]. While some studies focus on the
individual level of causality [25,30], others focus on group
level dynamics [2]; and Nelson, 1996 [56]; or the interorganizational or institutional level [21,44].
One of the reasons that there is no consensus denition
of organizational disaster is that the literature includes so
many proximate concepts. For example, the literature
includes studies of organizational failure (see [50] for an
overview), organizational catastrophe [5,48,60] and,
especially, organizational crisis [24,32,51,68]. Table 1
gives one a sense of the diversity of denitions of disaster
and crisis.
To avoid the problem of a missing or ambiguous denition, I posit that an organizational disaster occurs (for
a given organization) if highly disruptive events bring
extremely negative consequences to the organization or its
stakeholders (for elaboration see [12]). While it is sometimes not an easy matter to determine just what events do
and do not qualify as disaster, sufce it to say, the BP Gulf
oil spill is certainly a disaster by any standard.
3. Technology-embedded organizational disasters
A technology-embedded organizational disaster is
simply one in which technology plays an important part.
This does not imply that technology is the major determinant of the disaster; only that technology is signicantly

Table 1
Denitions of organization disaster and related concepts.
Organizational concept

Denition

Crisis

An organizational crisis (1) threatens high-priority values of the organization, (2) presents a restricted amount
of time in which a response can be made, and 3) is unexpected or unanticipated by the organization. ([33]; 64).
.We dene crises as events characterized by threat, surprise, and magnitude, with a need for a quick response
and high potential costs if they are not resolved effectively. ([28]; 27).
Crises are characterized by low probability/high consequence events that threaten the most fundamental goals
of an organization. Because of their low probability, these events defy interpretations and impose severe demands
on sensemaking. ([72]; 629).
Crisis implies a perception that an individual or set of individuals faces a potentially negative outcome. Crisis can be seen
as a composite perception based on several different dimensions of an issue. In particular, the perceived importance,
immediacy, and uncertainty of an issue all contribute to how threatening an issue is perceived to be. ([23]; 502).
Disaster is a type of routine nonconformity that signicantly departs from normative experience for a particular time
and place. It is a physical, cultural, and emotional event incurring social loss, often possessing a dramatic quality that
damages the fabric of social life. For an accident to be dened as a disaster, the accident would need to be large-scale,
unusually costly, unusually public, unusually unexpected, or some combination. ([70]; 292).
Industrial crises are disasters caused by human agencies and the social order; natural disasters are acts of nature.
The impacts of industrial crises sometimes transcend geographic boundaries and can even have trans-generational
effects. ([62]; 287)
Catastrophe refers to trauma in the workplace, typically arising from injuries, accidents or violence [5,43].
Failure, in organizations and elsewhere, is deviation from expected and desired results. This includes both avoidable
errors and the unavoidable negative outcomes of experiments and risk taking. ([18]; 300).

Disaster

Catastrophe
Failure

246

B. Bozeman / Technology in Society 33 (2011) 244252

implicated. This is in contrast to organizational disasters


owing to such factors as civil disturbance, economic
collapse, weather events, or the corrupt or immoral acts of
particular human beings.
4. Organizational and technology roles in disasters
In some cases where technology is implicated in
disaster, organizational factors play no prominent role
(even though organizations subsequently suffer great
damage). One good example is the 2002 explosion of the
Concorde jet taking off from Charles De Gaulle Airport [16].
This example is one of limited human agency. A piece from
another airplane was left on the runway, a Concorde tire
burst, materials were sucked into the engines, and the
plane was destroyed. This disaster could not reasonably be
said to have been precipitated by the pilots and crew of the
plane or by the company, Air France, responsible for the
well being of the passengers. No one made decisions or
took actions pertinent to the tragedy. Such disasters as
these, important and terrible though they may be, provide
little insight for the study of organizational disaster.
In other cases there is a clear human element in the
chain of events leading to the disaster. In such instances,
a point comes where parties disagree about risk and
procedure. In the case of the Challenger accident [70], that
moment came in discussions about the integrity of
components exposed to cold weather. A joint seal failed
due to the cold, precipitating a gas leak through the solid
rocket motor and penetrating the external tank, causing the
explosion [70]. But before the launch, there were discussions about the effect of the cold and whether to launch.
Ultimately, decision-makers interacting with each other
contributed considerably to the disaster.
In the case of the BP Gulf oil spill disaster, it seems clear
that human decision-making and organizational culture
played a key role. Human decision contributed to and
exacerbated the disaster [52]. However, as we shall see in
the case analysis below, the causal elements in the BP Gulf
oil spill are many, diverse and interdependent.
5. Theorizing about technology-embedded
organizational disasters
One approach to developing theory about organizational disaster is to consider the differences in their causes
as suggested by Fig. 1. Thus, we might expect a very
different approach to analysis when considering an HT/HH
type- the BP oil spill- than with the three other archetypes.
In considering each of the four quadrants, we can make
a few simple observations that perhaps clarify theorybuilding requirements. The lower right hand quadrant
(LH/LT) can be viewed as, essentially, acts of God, and
though they may be interesting from the standpoint of
emergency preparedness, they cannot generally be much
edied by organizational analysis. Similarly, the cases that
are driven almost entirely by seemingly unpreventable
technological failures, such as the Concorde disaster, have
limited interests to organizational theorists. In those cases
where disasters take on the aspect of stochastic technology
failures, unpredictable and largely unpreventable, the

disaster highlights needs for new technology and for


innovation, but has less relevance or use for organizational
theorizing. However, as one moves away from the extremes
to a role of human agency (e.g. Hurricane Katrina ooding),
there is a greater role to be played by organizational and
institutional analysis [11].
The predominant role for organization and institutional
theory is in the two HH quadrants (HH/LT, HH/HT); in such
cases, organizational and institutional theory may have the
ability to help explain disastrous outcomes and to prescribe
approaches to mitigation. In the case of the BP oil spill
disaster, there are likely important organizational and
strategic issues related to such factors as (1) coordination of
a multiplicity of contractors, (2) the relationship of organizational structures to human judgment interactions with
risk and failsafe technologies; and (3) the deployment of
organizational strategies in cases where extreme physical
risk meets nancial risk (i.e. contrasting, for example,
decision-making in economically sheltered nonmarket
organizations with similar decision requirements in highly
competitive market organizations).
Arguably, the organizations environments play an
important role in understanding an organizations role in
disasters and disaster mitigation. Fig. 2 adds two additional
dimensions, one related to environmental complexity and
the other to the degree of external constraint. As the gure
implies, the concepts presented here could potentially be
operationalized and tested against a larger set of organizational disasters.
6. BP oil spill as a case of technology-embedded
disaster
In the interest of space, I present the events of the BP oil
spill in a brief summary. This limitation is important to note
because the events surrounding the oil spill were exceedingly complex. Those wishing greater detail may wish to
consult the [52] Final Report.
We can have a better grasp of the various problems
encountered in the case if we understand what occurs
routinely in deep sea oil drilling. This enables us to better
identify and compare the problems experienced in this
exceptional case. In characterizing the operations of Deepwater Horizon, I rely extensively on the [52] nal report.
7. Deep sea drilling technology: what is supposed to
happen
Before April 20, 2010, the technological and management activities occurring on Deepwater Horizon (hereafter
Deepwater) were for the most part typical of contemporary deep sea drilling. The Deepwater rig was a prospecting
rig and would have been replaced by a smaller and less
expensive operation once oil reservoirs were conrmed
and the drill apparatus had been set up for suitable
extraction. Deepwater was designed to drill in very deep
water and at ocean oor depths of several miles. In 2009, it
had drilled the deepest oil well in history, more than
10,000 m, at a location in the Gulf of Mexico, 250 miles
from Houston, Texas [67].

B. Bozeman / Technology in Society 33 (2011) 244252

247

High Human Agency

(HH/LT)

(HT/HH)
Corps of Engineers
Everglades Dredging
BP Oil Spill
High
Technological
Culpability

Challenger

IRS Philadelphia Service


Center
Disaster
Low
Stanford Yacht Technological
Culpability
Scandal
Hurricane
Katrina
Flooding

1937 Texas Gas


Main Explosion

Concorde

(HT/LH)

Enron

2004 Indian Ocean


Tsunami

Low Human Agency

LH/LT

Fig. 1. Technology-embedded disasters.

The Deepwater was gigantic in size because of the need


to house scores of workers as well as a huge quantity of
supplies and equipment, especially drill pipes and associated technology. Deepwater could be thought of as not so
much a drilling apparatus per se but a small fabrication and
manufacturing facility.
After setting up, Deepwater, or any similar facility,
begins drilling in search of the oil reservoir. An important
part of the drilling, even at initial stages, is the injection of
so-called drilling mud. Drilling mud is a synthetic uid

lubricant, usually composed of bentonite clays or synthetic


compounds, used to compensate for the enormous amount
of friction that occurs with drilling [71]. In addition to
mitigating friction and the heat build up from friction,
drilling mud also serves as a conduit for material being
drilled, suspending materials such as sand or other
minerals and carrying materials to the surface [47].
Just as important as its role in inhibiting friction, drilling
mud is crucial for controlling the pressure that builds up
once oil or gas begins to rise. The drilling mud plays a role

High Human Agency

(HT/HH)

Environmental
Complexity

(HH/LT)

External Constraint

High
Technological
Culpability

(HT/LH)

Low
Technological
Culpability

Low Human Agency

LH/LT

Fig. 2. Technology-embedded disasters and organizational environment characteristics.

248

B. Bozeman / Technology in Society 33 (2011) 244252

throughout the drilling process as the mud is re-circulated,


passing through screens that trap materials from the
borehole and then cycling back to the drill bit. In general,
use of drilling mud is a key element of drilling effectiveness
and safety, and petroleum engineers strive to optimize the
amount, density and viscosity of the mud, making adjustments as needed by changing conditions and materials
ow.
Once an oil deposit is reached, the focus shifts to control.
Since oil is under heat and pressure, there is no need to
pump it, but rather to regulate its upward ow and,
particularly, to prevent an uncontrolled blowout. A blowout
is an unintended and uncontrolled release of crude from an
oil well that occurs after all the systems pressure control
devices have failed [36].
The heavy downward ow of the drilling mud is a major
tool for regulating pressure and preventing blowouts, as
such, a primary means of safety assurance. After the drill
has reached its site, a steel casing is inserted down the
middle of the bore. Many pieces of steel casing, each piece
about 10 m long, are screwed together as the casing
descends. In the BP Gulf site, the drill had extended for
more than three miles and, therefore the steel casing was
lengthy. After the casing is completed, the next step, also
a safety and control measure, is to pour concrete down the
watertight casing. A nal pipe is put through the casing,
and this pipe, surrounded by concrete, actually extracts the
oil.
Another crucial safety element is the blowout
preventer. This is a device set on the ocean oor above the
borehole that serves as a last resort for cutting off the ow
in case of loss of control or other emergency. The blowout
preventer is a unique piece of equipment and is manufactured on the rig and then lowered into place. The blowout
preventer includes one or more valves and may include
such components as electric and hydraulic control lines and
accumulators, test valves, electronic control pods, and kill
switches, all in a massive support frame. The devise used
for Deepwater was more than 50 feet tall and weighed more
than 300 tons.
Once all is in place, the shaft that the drill bit is connected to goes down to the oil deposit, and it is surrounded on the ocean oor by the blowout preventer. The
riser shaft extends up to the surface. The drill and the
drilling mud blow through the riser shaft and, ultimately,
the oil ows up, through the riser shaft and through the
blowout preventer, up to the rig. The pipes often are not
vertical but may be diagonal, depending on the most
efcient means of reaching the deposit. The blowout
preventer includes several different technologies for
sealing the drill hole, most of them intended for use only
in emergencies, with some used to more or less routinely
regulate volatile methane burps that rise through the
pipe. Events that, if not properly controlled, can easily
lead to explosions.
If all goes as planned, the conguration described
above is set in place and the oil deposits are released and
ow to the surface at a controlled rate, accumulating in
tanks at the rig. This extraction may go on for many
months, and, once all or most of the oil is extracted, the
well is capped.

8. The Deep water Horizon accident: what is NOT


supposed to happen- but did
In all likelihood, the BP Gulf oil spill would not have
occurred had the blowout preventer performed up to
specications. According to BPs own investigation (BP, Inc.,
2010, p. 9), the accident involved a well integrity failure,
followed by a loss of hydrostatic control of the well. This
was followed by a failure to control the ow from the well
with the BOP (blowout preventer) equipment, which
allowed the release and subsequent ignition of hydrocarbons. Ultimately, the BOP emergency functions failed to
seal the well after the initial explosions.
In most cases, the blowout preventer has several
different means of activation, some automatic. When activated, the blowout preventer is supposed to cut the drill
line and seal the well. For some reason, still unknown, this
did not occur in the Deepwater accident. BP ofcials suggested that a hydraulic leak could have been responsible,
but it is also possible that the emergency disconnect was
disabled by an explosion [26]. The valves were subsequently determined to have been shut to some degree, but
not sufcient to stop the ow.
There is some discussion that the blowout preventer
effectiveness was diminished by the fact that it did not have
an acoustic switch, a remote-control device for deploying
the blowout preventer. While such devices are often used
in rigs, the U.S. government (unlike some other countries)
does not require this third line of defense, a measure
costing about $500,000. Not being required to do so, BP
chose not to install this third tier device. According to one
report [35], the Department of Interiors Minerals
Management Service considered requiring remote-control
of blowout preventers several years ago but companies
protested, citing increased costs and undemonstrated
effectiveness.
According to the BP report (2010), largely consistent
with the [52] subsequent ndings, a series of failures
occurred, precipitating the disaster. Among the most
important:
 The cement mixture appeared not to be adequate and
allowed hydrocarbons to enter the wellbore.
 The Transocean and BP crew incorrectly concluded that
the pressure tests were adequate and proceeded to
replace the expensive drilling mud with seawater (a
common practice when well integrity has been established and pressure is judged at safe levels).
 The crew did not recognize the inux of hydrocarbons in
the riser in time to take corrective action.
 The blowout preventer failed, likely due to explosions,
causing the mechanisms autoshear function to be
disabled.
The BP report (2010, p. 11) concludes that accident was
due to a complex and interlinked series of mechanical
failures, human judgments, engineering design, operational implementation and team interfaces [that] came
together to allow the initiation and escalation of the accident. While other analyses (EPA, 2010 [52]; depart from
the BP report in assigning priorities to causes and in

B. Bozeman / Technology in Society 33 (2011) 244252

judging culpability, there seems little doubt about the


accuracy of the above description of events leading to the
disaster.

249

error tendencies are highly sensitive to organizational


culture and context. As [70] notes, research rmly links the
social origin [of disasters] to political, competitive, regulatory, and cultural environments (293).

9. Analyzing organizations and disaster


In the remainder of this analysis, I consider the interaction of technology and organizational elements. A rst
issue is the size and scope of organizational roles in technology-based disasters (or, from another perspective,
disaster avoidance and remediation). To what extent are
organizational disasters related to rationally constructed
strategies as opposed to poorly understood and largely
idiosyncratic events? Are organizations really hostage to
uncontrollable events in a sort of disaster lottery [12], or
do organizations more often implode because organizations promote cultures that underbid risk? After this
conceptual sorting, I consider organizational characteristics
pertaining to strategy and, specically, what characteristics of organizations and their cultures determine the level
and type of technology-based risk they are willing to
sustain?
10. Idiosyncrasy and the possibilities for analyzing
strategy
Some historians (e.g. [42] argue that any approach to
aggregation or generalization from organizational disasters
does not do justice to the role of idiosyncrasy or the
overdetermined [15] nature of complex historical events.
An idiosyncratic perspective tells us that we can develop
deep understanding of events only if we consider unique
and non-reoccurring attributes of those events. By seeking
to aggregate these attributes and study patterns across
series of unique events, one loses much information and an
important type of understanding.
To be sure, focusing on unique aspects of events
provides insights that emerge in no other way, and,
certainly, the act of aggregating requires some loss of detail.
I suggest that the limitations of aggregation and pattern
seeking pose no insurmountable barriers to systematic
inquiry about organizational disasters. For example, if one
reads the investigation reports of airline disasters, one
inevitably nds that a series of usually improbable idiosyncratic events are strung together to produce a causal
sequence leading to tragic outcome. At the same time,
however, these events often have close relation to organizational and policy variables such as pilot training protocols
and engineering design standards. In the case of the BP oil
spill disaster, one can fully credit the role of idiosyncrasy
while, at the same time, looking for patterns. Thus, even
while recognizing the interaction of human decisions, such
as premature replacement of drilling mud with seawater,
with highly idiosyncratic events, such as the breakage of
a single shearing device, one can examine organizational
issues. In the BP oil spill case, one might ask if the system
redundancy built in to these complex technology-human
interactive systems sufces to prevent the single failure
that can lead to disaster [45,46,65].
As reconstructions of such disasters as the Challenger
[69] and Three Mile Island [40] have shown us, human

11. Organizational coupling and disaster propensity


According to [54,55]; a major factor in organizational
disasters is interactive complexity. Social systems, especially tightly coupled ones, are amenable to failure due to
design aws. Often designs are so complicated that no one
could reasonably expect to understand fully all their
interdependencies. Indeed, this seems the case with the BP
oil spill, and, moreover, the interactive complexity is
exacerbated by the number of different organizations and
different organizational cultures that are parties to deep
sea oil exploration.
I argue that interactive complexity is at least as great
a problem in loosely couple systems, and that such systems
are often much more vulnerable to design aws because
their designs are so often soft or even chaotic. Perrows
insight about normal accidents is that even if they occur
only rarely in highly reliable socio-technical systems, they
will occur. However, we can expect that organizations in
loosely coupled systems, systems with ambiguous controls
and high autonomy, including the organizational and
policy domains in which deep water drilling occurs, will
more often produce disasters than will the more structured
systems (e.g. nuclear regulatory regimes) examined by
Perrow. In some cases failure is due to organizational
routines and disasters will ow from these routines.
12. Disaster theory and the BP oil spill
While it is true that organizational disasters have
a highly complex causal trajectory, in part because of the
signicant role of idiosyncrasy [19], one can nonetheless
discern important patterns and learn from them. The
organizational and institutional factors at the center of the
BP oil spill case seem to have much in common with other
well known and carefully analyzed technology-embedded
organizational disasters. Some of the common features
include (1) the role of hierarchy in mediating decisionmaking; (2) a false sense of security produced by redundant systems; (3) the apparent suppression of dissent (in
the case of substituting seawater for drilling mud); (4) the
diffusion of risk.
While the organizational level features of the BP oil spill
disaster have much in common with other well studied
organization-centered disasters, the broader policy and
institutional features of the oil spill case differentiate it
from such familiar cases as the Challenger disaster or the
Bhopal disaster. In the oil spill case one could well argue
that the bigger picture lies in understanding such factors
as public policy for offshore drilling, energy policy writ
large, and international commerce. In my judgment, these
different sets of factors are actually quite closely intertwined. Thus, public policy affects organizational perceptions of risk and energy policy affects requirements for
technology.

250

B. Bozeman / Technology in Society 33 (2011) 244252

As one looks for patterns to connect cases, one must also


consider just what aspects of a case render it sufciently
different to elevate its importance for theory development
(as opposed to its inherent importance for disaster
prevention). In the case of the BP oil spill disaster, two
factors stand out. In the rst place, BP, as seems to be the
case with many of its competitors, remains steadfast in its
opposition to many environmental and safety regulations
that, while having signicant short term costs, would seem
to have the advantage of enhancing predictability and
leveling the technological playing eld. What characteristics of organizations and their cultures lead to visceral
reactions to regulations, including, most importantly, ones
that could perhaps have, in this case, perhaps saved BP $40
billion? Companies seem to, at the same time, revile and
seek to co-opt regulatory agencies responsible for safety
[4]. How does such a corporate ethos emerge?
A second factor that stands out in this case is the wide
diffusion of responsibility. While everyone agrees,
including BP, that the prime contractor is ultimately
responsible for outcomes, how is it possible to exercise
effective monitoring when working with another companys technology, leased from a third party, engineered by
a fourth? When engineering decisions, such as those pertaining to lubrication engineering, is outsourced to other
companies with other designs and cultures, how is it
possible to maintain sufcient cohesion and to develop the
instantaneous information ows needed for crisis decisionmaking? These problems are not unique to the BP case and
they suggest why a comparative case approach may be
edifying.
13. Conclusions
The technology-embedded disasters model provides an
elementary framework for thinking about organizational
disasters. Even in its current, limited form, it has some
utility for helping understand organizational disaster and
the threat of the disasters. The model is taxonomic rather
than predictive, but, as is the case with any taxonomy, it
suggests the variables to be employed as entry points to
understanding. In particular, the model points to possible
policy and management emphases. To the extent there is
human agency involved (including not just individuals
actions but also acts or corporate individuals such as rms
and governments), disasters tend to be more complex but
also potentially amenable to policy solutions [6]. While
acts of God are, of course, relevant to public policy (think
of the levy failure in New Orleans in the wake of Hurricane
Katrina), they are only partly predicable and controllable.
Even with good policies, the winds blow. However, in
policies where human agency is to a greater degree
implicated, policies and strategies matter. Thus, in the wake
of Three Mile Island, Enron and IRS information technology
management failures, specic policies were adopted that
addressed threats.
We can at least see some nascent propositions emerging
from the technology-embedded disaster model as presented in Fig. 2. The model suggests that we should expect
a higher threat of disaster in cases when there is a complex
interaction of technology and human agency such as the BP

Gulf oil spill. However, despite this level of environmental


complexity, there is a relatively relaxed set of external
constraints and oversights. In the model simply identifying
this dangerous cocktail- the conuence of complex and
risky technology, human agency and high degrees of
interdependence, and limited constraint and oversightshould serve as a warning sign [39]. When we add to this
mix a potential for cataclysmic damage and organizational
cultures encouraging high risk for high payoff, the 2010
regulatory regime for deep sea drilling seems perfectly
maladapted [59].
The current model of technology-embedded disasters
has, at best, value as a sign post, suggesting emphases. The
model requires a great deal more development and more
comparative analysis before it has any potential as
a predictive model. What next steps are needed for a more
useful model? One way to address this question is by
looking at the model in Fig. 1 and asking, where does one
look for solutions? In some cases, the solution set to
disasters is close to nil, at least from a practical standpoint.
The above-mentioned Concorde case (low human agency
culpability, moderate technological culpability) seems not
to have been caused by poor performing technologies or
faulty organizations. In the case of the Enron disaster [61],
the solution set seems straightforward (though immensely
challenging) in that it points to developing a means of
constraining human venality. These means include changes
in regulatory regimes but perhaps also incentives and
means of choosing and advancing corporate executives.
What about the BP oil spill? Arguably, the most challenging realm of potential disaster is one in which both
technological and organizational systems play a major role,
both separately and interactively. Going beyond the model
to possible implications, I suggest that any case so characterized is fraught with great risk, and the following safeguards are in order: (a) a higher degree of system
redundancy and failsafe mechanisms; (b) an appreciation
for the difference between normal risk assessment and
assessment of potentially catastrophic risk; (c) greater
oversight of inter-organizational relations and (d) attendant mechanisms for accountability. The BP oil spill case
gives us a situation with only moderate system redundancy
and fail safes, limited attention to catastrophic risk,
remarkably limited oversight or even coordination of interorganizational relations and limited, diffuse accountability.
One of the most important factors not yet incorporated
into the technology-embedded disaster model is the
impact of resource scarcity and resource value [20,57] in
framing disaster potential and mitigation strategies. It is
noteworthy that even in the wake of one of the greatest
environmental disasters known to humankind, one of the
primary concerns on the part of policy-makers and the
general public was a fear that deep sea, offshore drilling
would be curtailed or limited. Indeed, legal challenges were
successfully mounted to thwart the Obama administrations moratorium.
The institutional and policy environment surrounding
the BP oil spill suggests a possibility for extending and
improving the technology-embedded disaster model. A
more powerful model would take into account the specic
nature and value of the resources at stake, the resources

B. Bozeman / Technology in Society 33 (2011) 244252

motivating organizational action and the deployment of


technology. Doing so could provide a better understanding
of disasters and of the possibilities for mitigating them. For
example, such a model would suggest time limitations of
remedial approaches relying largely on more regulations
and regulatory oversight. A precondition to the effectiveness or regulatory alternatives may well be an organizational and institutional environment amenable to
regulation. When resources are perceived to be so scarce
and valuable that (for many at least) environmental
catastrophe takes a back seat to resource pursuit, then
a useful model of disaster should lead us to consider just
how and why rapacious organizational cultures emerge
[59]. One possibility is that they exist with the complicity of
those nominally charged with overseeing them and with
the tacit consent of a public willing to make nearly any
sacrice or take any risk to ensure the continued ow of the
resources that emerge from marriage of high risk technology and risk tolerant organizational culture.

References
[1] Adcroft A, Hallberg R, Dunne JP, Samuels BL, Galt JA, Barker CH,
Payton D. Simulations of underwater plumes of dissolved oil in the
Gulf of Mexico. Geophysical Review Letters 2010;75(5):490502.
[2] Aldag RJ, Fuller SR. Beyond asco: a reappraisal of the groupthink
phenomenon and a new model of group decision processes.
Psychological Bulletin 1993;113(3):53343.
[3] Allison G. Conceptual models and the Cuban missile crisis. American
Political Science Review 1969;63(3):689718.
[4] Barley SR. Corporations, democracy, and the public good. Journal of
Management Inquiry 2007;16(3):20115.
[5] Barton Laurence. Trauma in the aftermath of organizational catastrophe: the short- and long-term impact on employees and their
supervisors. Disaster Prevention and Management 1994;3(1):1826.
[6] Birkland T, Nath R. Business and political dimensions in disaster
management. Journal of Public Policy 2000;20(3):275304.
[9] Bosman J. No shortage of books on oil spill. July 20, 2010, downloaded from. New York Times, http://www.nytimes.com/2010/07/
21/books/21oilbooks.html; 2010.
[10] Bowman E, Kunreuther H. Post-Bhopal behaviour at a chemical
company. Journal of Management Studies 1988;25(3):387400.
[11] Bozeman B. All organizations are public. San Francisco: Jossey-Bass
Publishing; 1987.
[12] Bozeman B. Toward a theory of organizational implosion. American
Review of Public Administration 2011;41(2):11940.
[14] Bozeman B, Rogers JR. A churn model of scientic knowledge value:
Internet researchers as a knowledge value collective. Research
Policy 2002;31(5):76994.
[15] Bunzl M. Causal overdetermination. Journal of Philosophy 1979;
76(3):13450.
[16] Bureau dEnqutes et dAnalyses. Report on accident on 25 July 2000
at "La Patte doie" at Gonesse, http://www.bea.aero/en/enquetes/
concorde/concorde-en.htm; 2002.
[17] Cameron KS, Tschirhart M. Postindustrial environments and organizational effectiveness in colleges and universities. The Journal of
Higher Education 1992;63(1):87108.
[18] Cannon, Edmondson A. Failing to learn and learning to fail (intelligently): how great organizations put failure to work to innovate and
improve. Long Range Planning 2005;38(3):299319.
[19] Carroll JS. Incident reviews in high-hazard industries: sense making
and learning under ambiguity and accountability. Organization and
Environment 1995;9(2):17597.
[20] Casciaro T, Piskorski M. Power imbalance, mutual dependence, and
constraint absorption: a closer look at resource dependence theory.
Administrative Science Quarterly 2005;50(2):16799.
[21] Covaleski MA, Dirsmith MW. An institutional perspective on the
rise, social transformation, and fall of a university budget category.
Administrative Science Quarterly 1988;33(4):56287.
[22] Canvar B. Disaster on the horizon: high stakes, high risks, and the
story behind the deepwater well blowout. Chelsea Green
Publishing; 2010.

251

[23] Dutton JE. The processing of crisis and non-crisis strategic issues.
Journal of Management Studies 1986;3(4):50117.
[24] Fink SL, Beak J, Taddeo K. Organizational crisis and change. Journal
of Applied Behavioral Science 1971;7(1):1537.
[25] Fischhoff B. Individual behavior in organizational crises. Organization Environment 1989;3(3):17789.
[26] Fountain H, Wald Matthew L. BP says leak may be closer to a solution. on-line. New York Times, http://www.nytimes.com/2010/05/
13/us/13spill.html; 2010.
[27] Fruedenburg W, Gramling R. Blowout in the Gulf: the bp oil spill
disaster and the future of energy in America. Cambridge, MA: MIT
Press; 2010.
[28] Greening DW, Johnson RA. Do managers and strategies matter?
a study in crisis. Journal of Management Studies 1996;33(1):2552.
[29] Gruca TS, Nath D. Regulatory change, constraints on adaptation and
organizational failure: an empirical analysis of acute care hospitals.
Strategic Management Journal 1994;15(5):34563.
[30] Hall DT, Manseld R. Organizational and individual response to
external stress. Administrative Science Quarterly 1971;16(4):53347.
[32] Hedberg A, Hedberg BLT. Responding to crisis. Journal of Business
Administration 1978;9(1):11137.
[33] Hermann C. Some consequences of crisis which limit the viability of
organizations. Administrative Science Quarterly 1963;8(1):6182.
[34] Holland J. Blowout, Plain sight- bp Deepwater Horizon cover up in
the Gulf. Create Space Publishing; 2010.
[35] Gold R, Casselman B, Chazan G. Leaking oil well lacked safeguard
device (April 28, 2010) downloaded from. Wall Street Journal,
http://online.wsj.com/article/SB100014240527487044235045752
12031417936798.html; 2010.
[36] Grace RD, Cudd B, Carden RS, Shursen JL. Blowout and well control
handbook. New York: Elsevier Publishing; 2003.
[37] Janis I. Victims of groupthink. Boston: Houghton-Mifin; 1972.
[38] Jas P, Skelcher C. Performance decline and turnaround in public
organizations: a theoretical and empirical analysis. British Journal of
Management 2005;16(3):195210.
[39] Kasperson RE, Kasperson JX. The social amplication and attenuation of risk. Annals of the American Academy of Political and Social
Science 1996;54(1):95105.
[40] Kemeny J. Report of the Presidents commission on the accident at
Three Mile Island. Washington, DC: USGPO; 1980.
[41] Kim DJ. Falls from grace and lessons from failure: Daewoo and
Medison. Long Range Planning 2007;40(45):44664.
[42] Knapp P. Can social theory escape from history? Views of history in
social science. History and Theory, 23;1:3452.
[43] Knefel AM, Bryant C. Workplace as combat zone: reconceptualizing
occupational and organizational violence. Deviant Behavior 2004;
25(6):579601.
[44] Knight L. Network learning: exploring learning by interorganizational networks. Human Relations 2002;55(4):42754.
[45] Landau M, Stout R. To manage is not to control: or the folly of type II
errors. Public Administration Review 1979;2(39):14856.
[46] LaPorte T, Consolini PM. Working in practice but not in theory:
theoretical challenges of high reliability organizations. Journal of
Public Administration Research and Theory 1991;1(1):3449.
[47] Lehner P, Deans B. In: Deep water: the anatomy of a disaster, the fate of
the gulf, and ending our oil addiction. Experiment Publishing; 2010.
[48] McKelvey B. Avoiding complexity catastrophe in co-evolutionary
pockets: strategies for rugged landscapes. Organization Science
1999;10(3):294321.
[49] Maltrud M, Peacock S, Visbeck M. "On the possible long-term fate of
oil released in the Deepwater Horizon incident,". Environmental
Research Letters 2010;5(3):17.
[50] Mellahi K, Wilkinson A. Organizational failure: a critique of recent
research and a proposed integrative framework. International
Journal of Management Review 2004;5-6(1):2141.
[51] Milburn TW, Schuler RS, Watman KH. Organizational crisis. Part I:
denition and conceptualization. Human Relations 1983;36(12):
114160.
[52] National Commission on the BP Deepwater Horizon Oil Spill. 2011
Final Report, January, 2011.
[54] Perrow C. Normal accidents: living with high-risk technologies.
New York: Basic Books; 1984.
[55] Perrow C. Accidents in high risk systems. Technology Studies 1994;
1(1):119.
[56] Peterson RS, Owens PD, Tetlock PE, Fan ET, Martorana P. Group
dynamics in top management teams: groupthink, vigilance, and
alternative models of organizational failure and success. Organizational Behavior and Human Decision Processes 1998;73(2/3):272
305.

252

B. Bozeman / Technology in Society 33 (2011) 244252

[57] Pfeffer J, Salancik GR. The external control of organizations:


a resource dependence perspective. New York: Harper and Row;
1978.
[59] Plater Z. Learning from disasters: twenty-one years after the Exxon
Valdez oil spill, will reactions to the Deepwater Horizon blowout
nally address the systematic aws revealed in Alaska? Environmental Law Reporter 2010;40(4):391408.
[60] Powell TC. Shaken, but alive: organizational behavior in the wake of
catastrophic events. Organization & Environment 1991;5(4):271
91.
[61] Prentice R. Enron: a brief behavioral autopsy. American Business
Law Journal 2003;40(2):41744.
[62] Shrivastava P, Mitroff I, Miller D, Miclani A. Understanding industrial crises. Journal of Management Studies 1988;24(4):285303.
[63] Starbuck WH, Milliken FJ. Challenger: ne-tuning the odds until
something breaks. Journal of Management Studies 1988;25(2):319
40.
[64] Reed S, Fitzgerald A. In: Too deep: BP and the drilling race that took
it down. Bloomberg Press; 2011.
[65] Roberts K. Some characteristics of one type of high reliability
organization. Organization Science 1990;1(1):16076.

[66] Simms R, Keenan JP. Predictors of external whistleblowing: organizational and intrapersonal variables. Journal of Business Ethics
1998;17(4):41121.
[67] Transocean. "Transoceans ultra-deepwater semisubmersible rig
Deepwater Horizon drills worlds deepest oil and gas well". Press,
Sept. 2, 2009. http://www.deepwater.com/fw/main/IDeepwaterHorizon-i-Drills-Worlds-Deepest-Oil-and-Gas-Well-419C151.html.
Retrieved December 22, 2010.
[68] Wicks D. Institutionalized mindsets of invulnerability: differentiated institutional elds and the antecedents of organizational crisis.
Organization Studies 2001;22(4):65992.
[69] Vaughan D. The Challenger launch decision: risky technology, culture,
and deviance at NASA. Chicago: University of Chicago Press; 1996.
[70] Vaughan D. The dark side of organizations: mistake, misconduct,
and disaster. Annual Review of Sociology 1999;25:271305.
[71] Wada K. Riser drilling technology and mud water drilling technology. Japan Agency for Marine-Earth Science and Technology
(JAMSTEC), http://www.jamstec.go.jp/jamstec-e/newsletter/vol5/
vol5_2.html; 2008. Downloaded December 11, 2010.
[72] Weick KE. The collapse of sensemaking in organizations: the Mann
Gulch disaster. Administrative Science Quarterly 1993;38(3):62852.