I have become intrigued by the many similarities in the way complex
bureaucracies with risk management responsibilities respond to risk assessment
messages and handle disasters.
PPT 2 The
first two case studies, presented here two years ago, were NASA and the
Columbia Shuttle accident and FBI headquarters' multiple failures to respond
to field office concerns about the men who would become the September 11th
terrorists. At that time, I characterized the many parallel contexts,
structures, and responses as "disaster by management." As I watched Katrina
and Rita unfold, I was struck by the repetition of many of these elements, as
well as the addition of a few novel elements, which I'll summarize here.
PPT 3
"Disaster by Management" includes the following components. First, the
agencies involved are under public-sector managerialist pressures.
Managerialism is the belief that government bodies are necessarily less
efficient with their resources than private companies, because they are not
kept in check by the profit requirement. Managerialism attempts to impose
corporate-style
cost-cutting and accountability measures on the operations of government
agencies and may include reducing or outsourcing part of the portfolio of
tasks done by the agency. NASA was subject to deadline pressures and saw more
than 40% of its Shuttle budget and workforce cut during the decade before the
Columbia crash. The FBI was experiencing internal reviews ordered by the
Justice Department concerning allegations of excessive use of FISA warrants.
Second, managerialist pressures tend to shift managers' balance of attention
from the precautionary principle to concerns with the opportunity costs of
excessive regulation. For NASA, concerns with schedule pressure led to
unreceptiveness to the risks of the foam strike. For the FBI, distraction by
the highest levels of management led to a failure to communicate the imminent
risk of an al-Qaeda strike to lower levels of the headquarters hierarchy,
which led to a lack of context in which to situate risk messages coming in
from the field offices.
Third, O'Connor differentiates agencies central to capital accumulation and
those more focussed on legitimation (e.g., social welfare or theatrical
functions). Legitimization agencies tend to be political footballs, with
variable funding and marked shifts of focus from administration to
administration. This certainly characterizes NASA and, now, FEMA.
Fourth, risk assessment science is lower in a bureaucratic hierarchy than
risk management policy decision-making. As risk messages ascend, their impact
may be diluted by the multitude of other managerial concerns with which they
must compete and may eventually become too weak in impact to trigger
appropriate managerial response. So, the engineers at Canaveral were daunted
by the initial denial of their request for Defense imagery, and the field
agents in the FBI had to accept the decision of lower level Headquarters
officers not to open investigations.
Fifth, organizational theorists debate the existence of high-reliability
organizations, with some holding to normal accident theory. High reliability
organization theorists describe organizational structures and redundancies
that permit the safe operation of high risk/high consequence technologies,
where normal accident theorists hold that the very redundancies and failsafes
themselves add to the complexity that makes cascades of failure inevitable.
NASA had been the template of a high reliability organization before
Challenger, and the FBI, too, had a similar reputation before 9/11.
Sixth, both NASA and the FBI, with their geographically ornate structures,
proved susceptible to the internal friction of spatial and social distance in
the ability of risk messages to move up or down within their organizations.
PPT 4
Looking more closely at FEMA's situation, the managerialist ethos of the Bush
43 administration was that government should not so much do work, even well,
as simply make sure that necessary work gets done. To that end, Budget
Director Mitch Daniels announced in April 2001 that FEMA would be functionally
divested of as much of its work as could feasibly be done by outsourcing to
the efficient private sector, even though it had become a "model agency of
good governance" in the 1990s.
In the governmental reorganization that followed 9/11, it was announced in
November 2003 that FEMA would be demoted from Cabinet level status and placed
under the new Department of Homeland Security. This was followed by the
resignation of FEMA Director Joe Allbaugh and the unexamined promotion of
Michael D. Brown to the directorship. In 2003, FEMA's central mitigation and
preparation functions were hived off and put in a separate Office of
Preparedness, and FEMA was to focus more on response to terrorism than natural
hazards. Frustration with its demotion, division of its once integrated
functions, and mission drift led to great frustration among FEMA staff, who
began resigning in droves, which eroded the institutional memory of the
agency.
PPT 5 As
if this weren't bad enough, FEMA was subjected to budget cuts. The White
House requested cuts of 23% from FY 2003 to 2005, which were eased to 10% cuts
after Congressional review. Project Impact, its showcase disaster prevention
through mitigation initiative, was cut out in 2001. The Hazards Mitigation
Grants Program was cut in half in 2003. In the wake of the Hurricane Pam
training scenario of 2004, which modeled a Category 3 hurricane squarely
hitting New Orleans, FEMA put in for moneys to buy and pre-position
communications equipment in and around New Orleans in its FY 2005 budget
request. This request was reviewed by DHS and edited out of its FY 2005
budget request. DHS, alone among Federal agencies, has the right to
"reprogram" money from one of its agencies to another without Congressional
review and did so multiple times in the three years before Katrina and Rita
struck.
PPT 6
Further affecting the performance (and morale) of FEMA, the agency has long
had a history of cronyism since it was first created in 1979. In the Reagan
and Bush 41 administrations, it was referred to as the "turkey farm," where
cronies could be given sinecures for campaign services rendered. The ideology
of those adminstrations was skeptical of the value of societal welfare
agencies and expressed that skepticism by putting inexperienced pals in them.
The consequences of cronyism were made painfully public in the ineptitude with
which FEMA responded to Hurricane Andrew in 1992. The failures became the
focus of an embarrassing study by the National Association of Public
Administration. President Clinton appointed his crony, James Lee Witt, to the
helm of FEMA and tasked him with the implementation of the NAPA suggestions
for reform. Witt came to the position, unlike his predecessors, with a
background in disaster management at the Arkansas Office of Emergency
Services. The agency instituted a "learning organization" culture and was
widely respected by the end of the 1990s as a high reliability, innovative,
and effective organization with a wholistic approach to the management of
hazards and the disaster cycle.
Bush 43, after initially telling Witt that he planned to keep him on, chose
Texas campaign fundraiser, Joe Allbaugh, to head FEMA, and he appointed to
FEMA Counsel his college roommate, Mike Brown, fresh from a managerial fiasco
at the International Arabian Horse Association, which forced that
organization's merger with another. FEMA was back to the turkey farm. After
some unfocussed shakeups within FEMA, even Allbaugh found the agency's
demotion too much of a slap in the face and he resigned, leaving Brown in
charge.
PPT 7
So, as FEMA entered the 2005 hurricane season, it was staffed at the highest
levels by inexperienced managers, had seen the frustrations of the core
staffers translate into resignations and the coring out of institutional
memory, the erosion of the thick ties among FEMA staffers and their
counterparts in state and local agencies, and its re-integration into at least
two disconnected chains of command.
PPT 8
The agency had lost a lot of know-how and no longer had the direct and
unsupervised right of access to the President. It could be overridden by DHS,
which could have eroded the authoritativeness of its urgent attempts, with the
National Hurricane Center, to communicate the magnitude of the risks posed by
Hurricane Katrina to the President in the teleconferences of late August
recently released by AP. Perhaps combining with the managerialist dread of
the opportunity costs of mitigation, preparation, and response, this lack of
authority could help explain the seeming apathy of the President in those
videos, in which he is never seen asking questions or showing curiosity but
only making bland assurances that resources would be, somehow, provided.
PPT 9 In
many ways, these problems and contradictions reflect the legitimization focus
of FEMA. FEMA was to secure the American population against disaster through
planning, mitigation, preparation, and coördination of response and
recovery activities. In other words, it was a classic social welfare agency,
and its fate reflects the fate of many a social welfare agency in an era of
managerialism and skepticism about the proper rôle of government. It
became a political football, was unable to defend its turf, and became a
turkey farm waiting for the disaster that would further reïnforce the
incompetence of the public sector. No wonder competent staffers left in
droves and the skeleton crew was unable to get the urgency of its message
about Katrina taken seriously by decision-makers with the power to orchestrate
the salvation of the vulnerable population.
PPT 10
Indeed, the whole hazards community of researchers and practitioners feels
marginalized, as expressed in this cartoon by Ron Pudim of the Boulder Hazards
Center.
PPT 11
Even in the best-run agencies, there can be a tendency to normalize anomaly,
to think that things will continue working the way they have in the past.
This tendency is one reason offered by normal accident theorists for the
inevitability of disaster. Faced with a run of successful narrow escapes,
managers get inured to the norm and try to frame even extreme anomalies within
their normal frameworks, which creates a lag in effective response.
Amplifying the tendency to normalize anomalies is the erosion of experience
among senior FEMA staffers, the pressure of other managerialist concerns at
the most senior levels, the junior status of risk assessment science, and the
extreme expense, in this case, of mitigation and preparation measures. All of
these are substantial hurdles for risk messages to jump and the normalizaton
of anomaly makes these enfeebled risk messages even more difficult to hear.
PPT 12
I had thought my interests in risk communications and the uneven dance between
risk assessment science and risk management policy decision-making had drifted
far from geography. Re-examining the three "disaster by management" case
studies, it is apparent that geography is a camouflaged player in all this.
In all three cases, geographically complex hierarchies underscored the junior
status of risk assessment science. Not only were the risk assessors who first
realized the importance of the hazards junior in rank to their management
supervisors, but their junior status was amplified by their housing in
spatially peripheral offices. The first concerned FBI agents were in the
field offices; the first concerned NASA engineers were at Canaveral and the
first concerned Boeing engineers in Texas felt some inhibition against
bothering Boeing software developers in Huntington Beach, California. And the
first concerned FEMA staffers were in regional offices. Hierarchy is
constituted and expressed at least in part by the spatiality of an
organization, and the friction of distance adds to the friction of social
distance that risk messages must traverse.
PPT 13
The takeaway message is that the ideology of public sector managerialism can
redirect public agencies from "good government" tasks that only they can
really perform, chasing the unexamined mirage of corporate efficiency until
their core functions are undermined. The accidents and disasters that ensue,
while seeming to underscore the image of governmental inefficiency, lead to
unnecessary loss of human life. The precautionary principle does run up
againt diminishing returns, but excessive concern for the opportunity costs of
governmental regulations and programs can and often does compromise human
safety.
PPT
14 The balance between Type I and Type II errors, as any statistician
will say, is a choice between the unavoidable possibilities of mistakes on the
basis of the worst potential consequences of each. Public sector
managerialism and the anti-government ideology it often expresses offer the
spectacle of barely imaginable human loss, as managers in government agencies
set aside the precautionary principle under the force of political stresses.
The selection of the balance between
de minimis calculations and the
precautionary principle needs to be a publicly debated process well ahead of
need.
