Volume 4,
August 2004
www.psljournal.com/archives/papers/pfiesteria.cfm
Scientific
Consensus and Public Policy:
The
Case of Pfiesteria[1]
Darrin
W. Belousek*
Abstract. This paper examines normative and
political aspects of the peer review, scientific consensus and public policy
processes related to harmful algal blooms of Pfiesteria in estuarine waters of
North
Carolina
and Maryland
in the 1990s. After laying out a
brief science and policy case history, the tension between the scientific
consensus and public policy processes in this case is analyzed in terms of
conflicts between scientific norms, public values and political expediency. The relationship between scientific
consensus and public policy in general is then questioned in light of this
case.
1.
Introduction
In
1991 North
Carolina
State
University
research scientist Dr. JoAnn Burkholder and her colleagues announced the
discovery of a toxic dinoflagellate, a single-celled marine organism, which was
named Pfiesteria piscicida, and
concluded that it was responsible for major fish kills in estuarine waters of
North
Carolina. Their findings were published in Nature (Burkholder et al. 1992). She and some of her assistants also
reported experiencing various negative health effects from working with the
organism in the laboratory. In
trying to bring the importance of these findings to the attention of both
North
Carolina
state officials and scientific peers, Dr. Burkholder was met with both public
and professional skepticism and soon found herself entangled in the politics of
both the state bureaucracy and scientific community. This story was sensationalized in a
popular book (Barker 1997). In the
midst of this tangle of science and politics, no coherent state-level public
policy was formed regarding either the human health hazard posed by or the
environmental control of the organism.
After
the release of the book in the popular press had brought national attention to
the situation in North
Carolina,
the presence of Pfiesteria was found
in conjunction with a minor fish kill in the summer of 1997 in a
Maryland
tributary of the Chesapeake
Bay. Despite an ensuing media “feeding
frenzy”, instead of a similarly messy tangle of science and politics, this event
was followed by a fairly harmonious working relationship between
Maryland’s
state public health and environmental management agencies and environmental
science research community. The
subsequent public policy-making process in Maryland quickly yielded two
significant outcomes supported by scientific consensus—a policy to close rivers
in which toxic Pfiesteria is found to
be present in order to protect public health, and regulations mandating nutrient
management practice for agricultural land use in order to enhance environmental
control of the organism.
This
study investigates how science and politics interacted to achieve or fail to
achieve public policy supported by scientific consensus. The outcome of the interaction of
science and politics is shaped by multiple, and often competing, values. Some of these values originate within
science itself as a knowledge-seeking enterprise. Other pertinent values derive from the
exigencies of the political process.
Still other values are derived from neither science nor politics but
rather reflect the human social ends that the public expects both science and
politics to serve. The locus at
which these values come into conflict usually occurs where uncertainties in the
available scientific information bear upon competing interests. After developing briefly the respective
case histories of Pfiesteria in
North
Carolina
and Maryland,
we’ll proceed to analyze how such values functioned in the present case and then
consider whether public policy ought to be based upon scientific consensus in
the first place.
2.
Case Histories
2.1
The science-policy background: estuarine nutrient
pollution
Over
the past thirty or more years, the international estuarine science community has
developed a broad and deep consensus that increased loads of nutrients such as
nitrogen (N) and phosphorus (P) into rivers and harbors is responsible for the
pollution of many coastal waters around the globe, including the
Abemarle-Pamlico (North Carolina) and Chesapeake Bay estuaries, with multiple
measurable deleterious effects on local ecology and economy. In waters over-enriched by nutrients,
growth of phytoplankton increases (‘eutrophication’), and algae in nutrient
over-enriched waters can bloom into dense populations that can have negative
environmental and human health impacts, both direct effects to fish and humans
via harmful toxins and several indirect effects, including loss of aquatic
vegetation, depleted levels of dissolved oxygen and fish kills.[2] Because of its toxicity, Pfiesteria piscicida is classified as a
‘harmful algae’.[3]
These
deleterious effects of nutrient pollution comprise the primary public
justification for a continuing coordinated process of research and policy in
both the Albemarle-Pamlico and Chesapeake
Bay
estuaries that has altered the landscape of science and politics in the
restoration and protection of coastal waters. For the purposes of this paper, it will
be helpful to review briefly the history of environmental science and public
policy regarding nutrient pollution in Chesapeake
Bay,
with a particular focus on Maryland.
From
the late 1960s to early 1970s, the nutrient-reduction agenda of the
environmental science and management communities focused primarily on
controlling point-sources of P-inputs (e.g., municipal sewage treatment plants);
here there was general congruence between the two, with management perhaps even
somewhat out ahead of the scientific consensus. During the mid-1970s, in the wake of
Tropical Storm Agnes, the scientific consensus moved ahead of management to
emphasize the need to control N as well as P from point sources and to control
non-point sources of N (e.g., suburban and rural land run-off) as well as
point-sources; the management community lagged behind partly because of
technical and economic factors in point-source N-removal. From the late-1970s to the mid-1980s the
EPA’s Chesapeake Bay Study synthesized and propelled the scientific consensus on
nutrient reduction, especially regarding non-point sources, leaving a ten-year
lag between science and management. In the mid- to late-1980s, the management
community began to catch up with the scientific community following the Bay
Study, leading to Maryland’s
1985 phosphate ban and the multi-state 1987 Chesapeake Bay Agreement that called
for a 40% reduction in the bay by 2000.
By the early 1990s, the scientific and management communities were
beginning to converge on the nutrient-reduction agenda, particularly in regard
to controlling non-point sources.[4]
During
the period 1991–1997 the environmental advocacy, science, and management
communities were building a solid shared consensus that agricultural sources are
the major contributor to nutrient pollution in the Bay and, hence, were
converging rapidly on mandatory non-point source nutrient controls for
agricultural land use practices as the next phase in the nutrient reduction
agenda.[5] An advancing edge in the agricultural
research community was pointing toward the need to rethink the contribution of
agricultural land run-off to nutrient pollution and the need for mandatory
nutrient management practices.[6] Lagging well behind this advancing
consensus and agenda, the Maryland state agricultural research and management
communities, however, with few exceptions, held to its conventional line that
agriculture did not present a nutrient pollution problem and was largely unmoved
by new scientific evidence that voluntary implementation of standard soil and
water conservation measures and conservation tillage practices were ineffective
in controlling farmland run-off.[7]
2.2
North
Carolina
Within
the environmental science research community in North Carolina there was
considerable dispute over Pfiesteria
concerning two main issues in particular—viz., whether massive fish kills
occurring on the Neuse River were due chiefly to Pfiesteria or depleted oxygen; and,
whether Pfiesteria posed a
significant human health risk to those exposed to Pfiesteria in open water. The latter scientific dispute, because
of its public health implications, embroiled the scientific research community
(especially Dr. JoAnn Burkholder) in a highly politicized controversy aired
publicly in the media.
Pfiesteria
is by all accounts a rather bizarre, complicated organism, regarding its life
cycle, nutrition and toxic-stage behavior.[8]
This strangeness led to keen
interest among some and appropriate skepticism among others when it was first
reported to the scientific community in connection to estuarine fish kills. Dr. Burkholder’s early laboratory
research on fish bioassays seemed to show fairly conclusively that Pfiesteria in its toxic stage could
cause fish mortality at sufficiently high concentrations of the organism. Whether Pfiesteria is the singular primary cause
of fish kills in estuarine rivers, however, cannot be inferred with certainty
from such laboratory results because in the field there are always multiple
confounding factors. In connection
with the fish kills on North
Carolina’s
Neuse
River
in 1995, for instance, both high Pfiesteria concentrations and toxic Pfiesteria activity as well as low
dissolved oxygen levels were variously observed. Dr. Hans Paerl, of the University of
North Carolina Institute of Marine Science and Burkholder’s chief local
scientific rival, concedes that Pfiesteria may be a secondary factor as
an opportunistic organism that strikes already-stressed fish populations; but,
he claims, low dissolved oxygen is the primary stressor responsible for the fish
kills.[9] This puts him at odds with Dr.
Burkholder; and their scientific dispute eventually made its way onto the pages
of a prominent marine biology journal, with Burkholder unabashedly charging
Paerl with poor science and questionable professional ethics (Paerl et al. 1998, Burkholder et al. 1999, and Paerl et al. 1999).
This
ongoing dispute between Burkholder and Paerl gained public attention in a Pfiesteria research peer-reviewed grant
process administered by North Carolina Sea Grant for North Carolina Department
of Environmental Health and Natural Resources (DEHNR) in 1995–96.[10] In 1994, Burkholder appealed to Gov.
Hunt and other state officials to support her Pfiesteria research and persuaded North
Carolina state Health Director Ron Levine to set aside some $600,000 of
unearmarked government funds for Pfiesteria research. She claims she was not seeking direct
state funding for her research, but she understood that there was a tacit
agreement with Levine that all of the research funds would go to her. DEHNR, though, instead decided to
distribute the funds through North Carolina Sea Grant, which routinely
administers research funds for the state.
This meant that Burkholder would have to compete with other researchers
for the funds as part of a standard scientific peer-review process. She objected to this decision, claiming
that it would unnecessarily delay crucial research and result in the funding
being spread out among less qualified researchers inexperienced with Pfiesteria. Despite these objections, she submitted
a research proposal seeking 75% of the available funds. Her research group received 75% of the
funds she requested, which amounted to more than 50% of the total funds awarded
to four different research groups including hers. In particular, she lost out on the
funding of the key nutritional ecology study, which went to Paerl’s research
group. She objected to this
outcome, claiming that Paerl’s research proposal contained errors, that
irregularities had occurred in the peer-review process and that she had been
deliberately sidelined in Pfiesteria
research by state officials and (then) Sea Grant Director Dr. B.J.
Copeland. Consequently, in protest
she returned to the state 80% of the funds awarded to her group, keeping only
money funding projects already underway.
In the wake of all this, the Governor demanded an accounting of the peer
review process from Copeland, and both the North Carolina State University
Chancellor and the state legislature launched investigations into the
dispute. In the end, Burkholder
received the funding she had returned via a direct grant from DEHNR Secretary
Jonathan Howes, who publicly apologized for how his agency had treated her, and
Paerl was cleared by the legislative audit. In the political fallout, however, Sea
Grant Director Copeland was dismissed by the NCSU Chancellor even though the
executive-level investigation found no wrongdoing on his
part.
This
official controversy spilled over into the debate concerning whether exposure to
Pfiesteria in open waters posed any
serious human health risk. One of
the studies funded by the Sea Grant peer-review process was an epidemiological
study by Dr. David Griffith of Eastern Carolina University (Griffith et al. 1998). This study, which found no serious
public health threat posed by Pfiesteria, was challenged publicly by
Burkholder as scientifically inadequate when a preliminary report was released
in April 1997 (Sheehan 1997). When
the final report of both this study and a Maryland public health study appeared
in 1998 bearing conflicting conclusions, the dispute erupted again in the media,
with Burkholder siding publicly with the Maryland study.[11] Consequently, the Griffith study quickly
became mired in the murky politics of Pfiesteria in North Carolina, buried
beneath accusations by Griffith of risk exaggeration (Griffith 1999a, b) and
counter-accusations by Burkholder of poor science and questionable professional
ethics (Burkholder and Glasgow 1999; cf. Lewitus et al. 1999 and Oldach 1999). This dispute inevitably drew in the
public health officials responsible for making public policy decisions, such as
state Health Director Dr. Ron Levine, who found himself under fire from
Burkholder and environmental activists in 1996–97, accused of doing too little,
too late to protect public health from Pfiesteria (Ready and Sheehan 1997 and
Clabby 1997). In the end, Levine
resigned as state Health Director and two other state health officials lost
their jobs over the controversy.
2.3
Maryland
During
1996 and 1997, there were several reports of strange lesions on fish found in
rivers on Maryland’s Lower Eastern Shore.
Just prior to the fish kill in the Lower Pocomoke River in August 1997,
the state convened a Technical Advisory Committee (TAC) consisting of experts in
fish pathology, algal ecology and environmental toxicology to evaluate
scientific evidence linking fish lesions, water quality and Pfiesteria. The TAC initially concluded in its
Interim Report that there was no conclusive evidence linking the fish lesions to
Pfiesteria, which by appearances
could have had multiple causes, or Pfiesteria to specific forms of
pollution or human activities (Center for Environmental Science 1997a). The TAC reconvened a month later after
the fish kill to review new evidence.
By contrast, its second report concluded, despite caveats about
uncertainties, that Pfiesteria was
the most probable cause of the fish kill and the observed lesions and that
nutrient pollution from human activity was the most likely environmental factor
promoting the growth of Pfiesteria
(Center for Environmental Science 1997b).
Regarding the policy implications of these findings, whereas the initial
report recommended continuing with the current nutrient reduction policy as the
only solution, the second report advised that the current policy was likely
insufficient to improve water quality enough to significantly control Pfiesteria and its
effects.
Along
with the reports of fish lesions, anecdotes of watermen that work the waters
with the lesioned fish complaining of various strange neurological symptoms had
also been accumulating during the previous year. Facing a potential public health crisis
in the wake of the fish kill, state health officials directed a medical team to
conduct a field study of people complaining of symptoms. On the basis of their findings, the
medical team recommended to state health officials that the river be closed,
which the state did just prior to the Labor Day weekend. A follow-up, longer-term clinical study
of the same subjects studied in the field confirmed the initial findings of
acute neuropsychological dysfunction and concluded that such symptoms were
correlated with open-water exposure to Pfiesteria (Grattan et al. 1998). This study left open the question of a
causal explanation of the symptoms, leaving uncertain the degree of risk to
public health posed by Pfiesteria, a
point emphasized elsewhere by one of the study’s authors (Oldach 1999). Subsequently, the state Department of
Natural Resources in consultation with the state health department developed a
policy to close rivers to public use where recent toxic Pfiesteria activity is found to be
present, using observation of Pfiesteria-associated fish lesions as
the sole environmental indicator for such activity based on the TAC consensus
(Department of Natural Resources 1997).
In
September 1997, Maryland Governor Glendening appointed a Citizens Pfiesteria Action Committee to make
public policy recommendations about specific environmental regulations related
to controlling Pfiesteria prior to
the January 1998 legislative session.
While this commission was in process, Dr. Donald Boesch, President of the
University of Maryland Center for Environmental Science and ex officio member of
the Governor’s bay cabinet, after recommending to the chairman of the Commission
that the process needed to be informed by the best scientific information
available, convened a panel of scientists to review the available information
and draw a consensus on the linkage between Pfiesteria and nutrients. This scientific panel, the Cambridge
Forum, which included Dr. Burkholder and relied heavily on her research,
concluded after its meeting in October 1997 that, “In the long term, decreases
in nutrient loading will reduce eutrophication, thereby improving water quality,
and in this context will likely reduce the risk of toxic outbreaks of Pfiesteria-like dinoflagellates and
harmful algal blooms” (Center for Environmental Science 1997c). In other words, against the background
of the scientifically fortified and politically entrenched environmental
management agenda of nutrient reduction in Chesapeake Bay, the scientific
consensus made the effectively conservative policy recommendation, “more of the
same.” The Citizens Commission then
used this scientific consensus—in addition to the advancing edge of agricultural
research on nutrients and farmland run-off that, due to prevailing political
conditions, was forcing along the state agricultural community (College of
Agriculture and Natural Resources 1997)—as the basis for recommending a policy
mandating stringent nutrient management practices for nearly all agricultural
operations in the state (Hughes 1997), a policy that was enacted into law in
1998 as the Water Quality Improvement Act (WQIA), which is described as “the
most comprehensive farm nutrient control legislation in the country” (Maryland
Cooperative Extension 1998).
3.
Norms in Conflict
What we aim for in this section is not an “external” social-theoretical
analysis of this case in terms of abstract sociological theories and concepts,
but rather an “internal” analysis on the scientific community’s own terms: Where
did the scientific community fail to live up to its own norms in forming
consensus on policy-relevant questions?
And how, if at all, did those failures affect either the scientific
consensus or the public policy based upon scientific consensus? By initially orienting such questions
from the perspective of the scientific community, we do not intend to adopt a
naïve ‘realism’ about science and policy (see next section). Nor do we suppose that an “internal”
perspective by itself is, in the last analysis, sufficient for characterizing
both actual scientific practice and the actual relationship between scientific
consensus and public policy; indeed, a fully adequate characterization of
science-policy interaction requires as well analysis from an “external”
viewpoint, some aspects of which we will consider in the next section. Nonetheless, it would seem that a
comprehensive accounting and assessment of the relationship between science and
policy would be incomplete without an “internal” perspective; and taking
seriously the self-understanding of scientists regarding their own practice need
not presume a ‘positivist’ value-neutral epistemology, a matter to be taken up
in a further study.
3.1
Normal v. Mandated Science
Science
as it functions within the professional boundaries of the scientific research
community itself and science as it functions under mandate from government
officials/agencies can often be in considerable conflict because of the
differing values and standards used to judge science within these respective
contexts. ‘Normal science’ is that
everyday science that may be described as being framed professionally and
institutionally within a particular ‘paradigm’ of research or ‘research
program’. Merton (1942) articulated
the ‘ethos’ of ‘normal science’ in terms of four norms of practice:
universalism, communalism, disinterestedness and organized skepticism, which
characterize the egalitarianism, economics, politics and sensibility of the
scientific research community, respectively.[12] He did not understand these norms to be
sociological descriptions, factually descriptive characterizations universally
true of actual scientific practice; science, always and everywhere, surely does
not actually function according to these norms. Nor are these norms to be taken as mere
social conventions, representative of a power balance achieved by negotiation
between competing factions within the research community; thus, they are not to
be sacrificed whenever they interfere with individual advantage.[13] Instead, Merton understood these norms
as “institutional imperatives (mores)” that are derived from the collective goal
(viz., “extension of certified knowledge”) and the technical methods (empirical
evidence and experimentally confirmed laws) of science; together with the
technical standards of logical consistency and accurate prediction, he claimed,
these norms (ideally) promote progress toward the goal of science.[14]
For
the scientific research community, such norms do express ideals to which
scientists themselves for the most part subscribe, whether or not these are
always followed in actual practice.
From an “internal” perspective, these norms are primarily prescriptive,
defining patterns and standards of behavior which scientists expect their colleagues to follow. The source of the prescriptive force of
these norms is the research community itself, its aims, practices and
institutions; and the scientific community observes itself regarding these norms
(though imperfectly, for sure).
There has been much discussion recently, for example, about scientists
keeping data secret and patenting research techniques and products; and
scientists describe a trend of “decline” from the norm of communalism due to
“external” values such as pressure from corporate patrons and the lure of
financial profit (Marshall 1990).
Scientists themselves can observe an ethic to decline only once it has
already been (perhaps tacitly) accepted as prescriptively normative for their
community. Serious violation of
these norms can be met with unofficial sanctions of various forms: denial of
promotion, denial of funding, loss of position or pressure for resignation, loss
of editorial positions, refusal from colleagues to cooperate in research,
etc. These norms thus also function
effectively as social indicators.
Following these norms in one’s research signals to other scientists that
“you are one of us,” a member of the research community, ready and willing to
“play by the rules” of science; conversely, failure to follow these norms opens
a question in the mind of the community as to the authenticity of one’s
credentials as a scientist.
A
brief discussion of these norms will assist in the analysis to follow. The norm of universalism expresses the
idea that science as an institutionalized practice seeks knowledge characterized
by an objectivity that transcends race, nationality, religion, class, gender and
personal biases; there is to be no “particular” science, but one “universal”
science that is true (or false) independently of the particularities that
characterize individual practitioners of science and the research communities in
which they practice (even though scientists and research communities cannot
escape such particularities in actual practice). Communalism is the notion that in
science there is to be a “common ownership of goods” that the scientific
community produces, exhibited concretely in the practices of open sharing of
data and materials; there are to be no individual proprietary rights in science
except the right to recognition of claims of priority and originality and to
institutional rewards such as prizes and promotions (i.e., scientists are never
exclusive owners of their own discoveries). Disinterestedness, while not necessarily
a character trait of individual scientists (each scientist has his or her own
particular interests and motivations for doing science), signifies the idea that
science is to have collectively neither a particular “clientele” (as the legal
and medical professions do) nor a particular “patron” (as the arts might) whose
interests foreign to science can compete with or take priority over the
institutional goal of objective knowledge.
Unlike literature and the performing arts today existing in symbiosis
with professional critics, and unlike philosophy in medieval scholasticism being
prized as a “handmaiden to theology,” science is to be judge of itself as
science (i.e., as a knowledge-seeking enterprise); and such autonomous judgment
is to be realized through the multi-faceted activity of ‘peer review.’ Organized skepticism refers to the
collective attitude with which the scientific community filters knowledge claims
in order to winnow out a stable consensus widely shared within the community;
this involves the skeptical scrutiny of knowledge claims according to logical
and empirical criteria detached from implications of the moral value and social
utility of science. This scientific
sensibility, undisturbed by the siren calls of profits and fame and unharried by
the pressures of funding and politics, tends to follow (ideally) Francis Bacon’s
dictum that “truth is the daughter of time” rather than of authority or
tradition.
None
of these norms stands by itself in the practice of science: if the scientific
community fails to observe any one norm, then its ability to fulfill all the
others, and hence achieve its institutional goal, is compromised. They operate socially and
institutionally within what may be described as a ‘knowledge filter’ (Bauer
1992).[15] The knowledge filter functions over an
extended, undetermined period of time to take initial speculations, hunches and
ideas that are generally unreliable and subjective (influenced by human biases
and fallibility) and transform them into relatively objective and reliable
knowledge (viz., textbook science) via a series of (imperfect) ‘filtering’
mechanisms whereby “individual frailties or imperfections must run the gauntlet
of communal scrutiny” before hypotheses are accepted as knowledge: peer review
of research grant proposals, replication of research results in multiple trials,
comment by colleagues in research seminar presentations, comments by editors and
referees of professional journals, testing and use by other scientists that
modifies and extends the original work, etc.
‘Mandated science’ is science done for public policy-making purposes and
takes two forms: original investigations commissioned by government officials or
regulators, and reviews (either by an individual or by a consensus committee) of
peer-reviewed science done originally for ‘academic’ purposes (Salter
1988). Mandated science relies upon
the ideal image of normal science as objective knowledge. Indeed, the very turn to science by
government officials and agencies concerning public policy matters that bear on
competing private interests is based upon the assumption that science can be a
‘neutral arbiter’ (or, at least, an ‘honest broker’) between those
interests. The exigencies of the
political process, however, often make demands upon science that compromise the
norms of science. Chief among those
demands is time: when government mandates science to provide information to aid
in decision-making, science must conform to the time-cycle of politics if it is
to be relevant to public policy.
Such time constraints can result in science-based public policies that
reflect an uncertainty-laden, quickly-formed consensus of a mandated report
instead of the time-consuming, carefully-winnowed consensus of the ‘knowledge
filter.’ Not only the pressures of
politics but the social significance of the science can permit serious
distortion of normal science by the media and invite interference by public
officials, and the lure of public recognition in high profile cases (like this
one) can lead scientists to circumvent or abandon the accepted norms.
There
are several aspects in which this case exhibits significant tension between
normal and mandated science as well as failures of the research community to
observe the norms of scientific practice.
Here we will focus on three episodes in particular, the North Carolina
Sea Grant research funding process in 1995–96, the public policy-making process
in Maryland in 1997-98, and the Maryland and North Carolina consensus committee
processes in 1997.
3.2
Peer Review/Openness v. Politics/Interests
The
North Carolina Sea Grant funding process was marked throughout by departures
from the norms of disinterestedness, communalism and organized skepticism. The process began with a personal appeal
by Dr. Burkholder to Gov. Hunt for what would have been (by all appearances)
direct state funding for her research.
Now, while government funding of research does not of course necessarily
distort normal scientific practice, the standard mode by which the state’s
interest is mediated to the scientific community is the peer-review system so
that it is the scientific community that judges according to its own standards,
criteria, and values what or whose research is worth funding. When state officials decided to channel
the Pfiesteria funding through a
standard scientific peer-review process, Dr. Burkholder objected, claiming that
doing so would cause unnecessary delays and end up funding less qualified
researchers. When the peer-review
process did not fund all the research she had proposed, she not only rejected
the process as lacking integrity and withdrew, again seeking (and this time
obtaining) direct funding from the state, but subsequently she refused to openly
share data and materials (specifically, assistance with organism identification,
maps of Pfiesteria blooms, and Pfiesteria culture samples) with those
researchers whose proposed projects were funded. And as soon as preliminary reports of
some of those research projects were released by Sea Grant, she criticized that
research through the media rather than through the standard modes of scientific
communication. Such a
personality-driven, media-exposed affair in science invited politicization of
and interference with the peer-review process, which resulted in science being
subjected to the standards of public rationality and political
process.
All
this helps make sense of the (hyperbolic) remark by Dr. Joe Ramus of the Duke
University Marine Biology Laboratory that this chain of events amounted to “the
total failure of a principle of science.
The peer-review process has been circumvented...it has collapsed.”[16] From her own perspective, Dr. Burkholder
agrees that the peer-review process has fallen apart, but she locates the
problem with others in the scientific community whose criticism of her work, she
claims, “has gone beyond healthy scientific skepticism.”[17] Nonetheless, she candidly asserts that
the standard peer-review process shouldn’t even apply in this case, which in her
view presents a fundamentally different situation because, as she sees it, most
of her peers are simply unable to adequately assess her unique expertise in Pfiesteria research: scientific peers
who lack experience with Pfiesteria,
she says, can’t appreciate her and her collaborators’ competence to work on Pfiesteria and, hence, should not be
permitted to evaluate her research proposals.[18] This way of thinking carried over
directly into her decision not to supply cultures to her competitors for the Sea
Grant funding—viz., because in her eyes they lacked the requisite technical
competence and professional integrity.
An additional, non-trivial issue in the sharing of Pfiesteria cultures is the time and
expense involved in producing and maintaining the cultures. Dr. Burkholder (understandably) does not
want her laboratory to be reduced to a culture-producing factory serving other
researchers (which, she says, would happen if she directed the efforts of her
assistants to fulfilling every request for cultures). She is clear to say that if given
funding for culture production, she is willing to share her cultures with other
researchers, subject (she emphasizes) to her private assessment of their
competence and character.[19]
Dr.
Burkholder’s decision to provide data, materials, and cooperation on a
qualified, private-assessment-only basis to scientists whose research proposals
had been approved by a standard peer-review process raises the question of who
decides who is worthy and competent to perform research. Her response to the peer-review process
shifted that question concerning her own research from the arena of scientific
evaluation to the arena of politics and advocacy, in which she enjoyed
“patronage” relationships (viz., favor from both the Governor’s office and
environmental groups). She
rationalized this response by a claim of unique qualification. This in effect created an exception for
her and her research from the standard practice of the research community, which
runs contrary to the accepted scientific norm of open sharing of data and
materials. This norm has been
recently re-articulated and re-affirmed unequivocally by the National Research
Council:
…[M]ost
arguments for making exceptions to standards could not be rationalized without
sacrificing the integrity of the principles of publication…[E]xceptions unfairly
penalize the community, which would have otherwise had access to the data,
information or material being withheld.
Furthermore, granting a special exception to certain categories or
particular researchers is problematic for a variety of reasons, including the
difficulty of deciding who qualifies for the exception…Universal adherence,
without exception, to a principle of full disclosure and unrestricted access to
data and materials that are central or integral to published findings will
promote cooperation and prevent divisiveness in the scientific community,
maintain the value and prestige of publication, and promote the progress of
science. (National Research Council
2003, pp. 13-14)
It thus seems plausible to conclude here that there occurred in this case
a significant compromise of normal science in the midst of a public-policy
relevant scientific dispute that pitted political and bureaucratic values (viz.,
efficiency of process and inscrutability of expertise) and techniques of
advocacy (e.g., use of media and exposure of opponents) against the scientific
norms of independence of peer review and openness of data and sharing of
materials. This compromise was
engendered by the elevation of one scientist’s expertise above peer criticism
and, correspondingly, (at least a temporary) exemption of one scientist’s
research from the skeptical scrutiny of the locally relevant research
community. The resulting
dysfunction of the research community on its own terms consequently undermined
potential for a scientific consensus unmixed with local politics that could
inform policy with at least some veneer of objectivity that would be convincing
to the public (see below).
3.3
Skepticism v. Time
In
the Maryland public policy-making process in 1997, one sees a tension between
the normal scientific sensibility of organized skepticism and the time-pressures
of politics. Although the initial
TAC consensus confirming the Pfiesteria-fish lesion linkage was
hedged with caveats about uncertainty, politics could not wait for
pathology. At stake were certain
public values—including cultural,[20]
economic[21]
and, as already mentioned, public health factors—that, compounded by media
attention[22]
and unfounded public fear[23]
that were dubbed “Pfiesteria
hysteria” (Greer and Leffler 1997), put the policy-making process under pressure
to produce politically effective results.
The chief question here is not what political motivations beyond concern for public
health and protection of natural resources Governor Glendening might have had
(say, re-election in 1998) in appointing the Pfiesteria Commission and setting its
timeline for action, or whether such political motivations may have conflicted
with the epistemological pursuits of the scientific research community. The primary issue here is the structural-institutional mismatch
between science and politics—viz., that public policy-making inevitably conforms
to legislative patterns and election cycles that necessitate timely compromise
accountable to public values, while normal science (except indirectly via
periodic funding requests) does not tend to do so.
Thus,
Maryland’s Governor-appointed Citizens Pfiesteria Action Commission could not
wait in the fall of 1997 for the lesion question to be resolved by further
research prior to addressing the more legislatively salient issues of nutrient
pollution and land use regulation in time for the start of the legislative
session in January 1998. In meeting
its mandated time constraint, the Citizens Commission, confident in the early
TAC consensus, by-passed the lesion question and went straight to the nutrients
issue. Meanwhile, laboratory fish
pathology investigations were conducted on the etiology of the fish lesions
associated with Pfiesteria blooms in
Maryland’s Lower Eastern Shore rivers, with preliminary results in 1998 showing
multiple possible causes (Blazer et
al. 1998 and Kane et al.
1998). Thus, during the course of
the year since the TAC concluded that Pfiesteria was the most probable cause
of the lesions, the scientific uncertainty over lesion etiology was only
magnified by the organized skepticism of the research community.[24] As a result of the research results
released in 1998, the TAC subsequently vacillated on whether Pfiesteria was even a primary cause of
the lesions—“Pfiesteria
toxins...cannot be ruled in or out as initiators of fresh lesions or deep
ulcers” (Center for Environmental Science 1999)—and Maryland Department of
Natural Resources officials revised the river closure policy
accordingly.
Guided
by the early TAC consensus on lesions based on available information, the public
policy-making process eagerly leapt ahead in the chain of causation to those
linkages—Pfiesteria-nutrients and
nutrients-land use—having the greatest political currency for leveraging change
in current agricultural nutrient control policy. Consequently, the scientists of the
Cambridge Forum, convened in order to inform the public policy-making process,
operated under the constraints of the Citizens Commission agenda—viz., to reach
consensus solely on the Pfiesteria-nutrients linkage— and
therefore effectively bracketed the Pfiesteria-lesions linkage, which was
presumed by the Commission. As
further scientific evidence on lesion etiology became available, the TAC
appropriately revised its consensus to loosen the Pfiesteria-lesion linkage, but too late
to have any effect on a nutrient management policy-making process that had
already been completed several months earlier. Now, even had the public policy-making
process waited on the peer-review process, it cannot be said with certainty what
impact this might have had on the politics of Pfiesteria. As the TAC was clear to point out in its
latter report, regardless of lesion etiology, Pfiesteria remains both a fish and human
health concern (Center for Environmental Science 1999). Nonetheless, the effect of this time-lag
between mandated and normal science was that environmental interests were able
to use the early consensus as leverage for their political agenda of stronger
nutrient control, playing public resource values against private property
interests, a strategy that would surely have been confounded by the later
consensus on lesions despite the TAC caveat regarding public health.[25] Moreover, considering sentiment among
Lower Eastern Shore voters just prior to the 1998 election and the reaction of
one member of the state Agricultural Advisory Board to the further research on
lesions (Smith 1998 and Shelsby 2000), it seems safe to say that agricultural
interests would have played up the increasing scientific uncertainty as calling
into question the whole of the Pfiesteria hypothesis as an insufficient
scientific basis for the nutrient control legislation.
Thus
we see that the mismatch in time-scale between the public policy-making process
and the peer-review process likely benefited those political interests in favor
of stronger regulation of agricultural land use. Because of this mismatch, mandated
science in this case was effectively biased toward environmental interests, and
thus the ideal image of normal science as objective was compromised. Does this imply that the scientists
involved failed to be ‘honest brokers’ of information? Answering that question may depend upon
what type of risk assessment strategy one thinks scientists should use to
evaluate hypotheses in the face of uncertainties, a topic beyond the scope of
this paper.[26] For sure, though, the upshot here is
that mandated science cannot always achieve a timely consensus on new and
limited information that is neutral between competing political
interests.
3.4
Universality v. Personality
The
norm of universalism was seriously compromised in the consensus committee
processes in North Carolina and Maryland, which suffered from personality
conflicts within the scientific community stemming from the public controversy
in North Carolina. The norm of
universalism regarding scientific consensus concerns chiefly who is, and who is
not, invited to the consensus table, and who decides. [27] In this case, Dr. Burkholder was in
effect given power to determine (at least in part) who was and who was not
invited to both the TAC and the Cambridge Forum; and that power was granted (in
part, at least) for sake of political expediency.
Dr.
Burkholder was invited to serve on the TAC at the request of Maryland State
Secretary of Natural Resources John R. Griffin, from whose department the TAC
received its mandate. Dr.
Burkholder initially hesitated to attend the first TAC meeting, however, because
public health officials from North Carolina had also been invited. She agreed to attend only on the
condition that they not attend; and so they were uninvited.[28] Thus, the political situation in North
Carolina effectively limited the scientific consensus in
Maryland.
The
convening of the Cambridge Forum presents a similar scenario. After the TAC’s second meeting, Gov.
Glendening appointed the Citizens Pfiesteria Action Commission. The Cambridge Forum, while not
officially mandated by the Pfiesteria
Commission, was convened by Dr. Boesch for the purpose of informing the
policy-making process. Dr. Burkholder gave testimony by invitation at the
Commission’s first meeting and then testified on Pfiesteria before a U.S. House of
Representatives subcommittee just a few days later. By this point, she was the recognized expert on Pfiesteria as far as the Commission was
concerned; hence, if the Cambridge Consensus were to carry weight with the
Commission, Dr. Burkholder would have to be included. Above politics, scientific expertise was
also at stake; at the time, Dr. Burkholder had authored most of the
peer-reviewed published literature on Pfiesteria. Now, Dr. Boesch had initially wanted to
invite both Dr. Burkholder and her North Carolina rivals Dr. Paerl and Dr. Pat
Tester to be on the Forum panel, even though he was well aware from the book And the Waters Turned to Blood that
there were serious tensions between them.
Dr. Burkholder, however, balked: it was either her or them; she would not
come if they were invited. So, it
would have to be Dr. Burkholder.
Again, to be fair, it should be said that Dr. Boesch’s decision itself
was based at least as much on the question of scientific expertise as on
politics; Dr. Paerl’s research was as yet unfinished and unpublished (Paerl and
Pinckney 1998).[29] Nonetheless, that there was an
“either-or” choice to be made was a matter of personality conflict stemming from
the political situation in North Carolina; and that conflict effectively
diminished the universality of the scientific consensus committee. According to Dr. Thomas Malone, member
of the Cambridge Forum, there was not an optimum mix of scientists on the
Cambridge Forum, because of both time constraints on drawing the committee
together as well as the impact of Dr. Burkholder on who could and couldn’t be on
the committee.[30]
Similar
problems were faced in North Carolina by those who in December 1997 convened the
scientific panel that produced the Raleigh Report (Water Resources Research
Institute 1998). The responsibility
for convening the panel fell to Marion Smith, then assistant to the North
Carolina Secretary of Environment and Natural Resources, and Dr. Kenneth
Reckhow, Director of the Water Resources Research Institute at North Carolina
State University. In Smith’s view,
dissension between Dr. Burkholder and other North Carolina scientists was
inhibiting effective influence of science on policy; thus there was a need for
independent evaluation of Pfiesteria
research to help the policy process along.[31] When Dr. Reckhow initially contacted Dr.
Burkholder to discuss what areas of expertise should be represented on the
panel, which scientists should be invited, and which research papers should be
consulted, however, he soon found himself mired in just the personal
controversies and animosities they wanted to avoid.[32] Subsequently, to rise above the politics
of personality, Smith and Dr. Reckhow decided not to invite any researchers from North Carolina;
neither Dr. Burkholder nor Dr. Paerl were invited, nor were any of their close
collaborators. Instead, they
gathered an impressive array of both government and university researchers from
both the eastern U.S. coast and Canada.
Those scientists most intimate with Pfiesteria research and closest to the
public policy-making process in North Carolina, however, were necessarily
excluded because of ties to Burkholder or Paerl. As a result, the Raleigh Report proved
to have little impact on water quality policy in North Carolina. Thus, whereas avoiding personality
conflict in North Carolina perhaps allowed for a scientific panel representing a
broader national-international cross-section of the research community (and,
hence, greater universality), doing so necessarily diminished the local
expertise present at the consensus table (and, consequently, the local political
relevance of the consensus).
Thus,
in the two cases one sees an anti-correlation between the universality of the
mandated scientific consensus and the political relevance of that
consensus. In Maryland, ensuring
the political relevance of the scientific consensus by including Dr. Burkholder
came at the price of excluding scientific colleagues who would have broadened
the universality of the consensus.
In North Carolina, the attempt to avoid local politics resulted in a more
universal consensus that was relatively politically irrelevant because the state
scientific community had little stake in it. Did this tension between normal and
mandated science undermine the objectivity of the scientific consensus? While that is difficult to assess, one
can point to the fact that the Cambridge Consensus and the Raleigh Report reach
virtually the same conclusion on the Pfiesteria-nutrients linkage as evidence
in favor of the scientific community having filtered political geography out of
its consensus in this case. Insofar
as the scientific community was able to do so, it was likely due to its reliance
in both processes not so much on the specific expertise of any one particular
scientist regarding Pfiesteria but
rather on the prior shared consensus
regarding nutrients, eutrophication and algal blooms, which was developed
independently of, and hence was largely immune to, the twinned politics of
personalities and Pfiesteria.
4.
Scientific Consensus and Public Policy
4.1.
Science and policy—naïve ‘realism’
One view of the relationship of science and policy (often considered to
be naïve) may be called ‘realism.’
The term ‘realism’ here is taken from Jasanoff’s terminology concerning
the relationship between science and policy and refers to the view that
truths
about the natural world arise without meaningful human agency or intervention,
in an autonomous domain of endeavor that is cleanly separated from the uses of
political power. Facts, the results
of scientific inquiry, are assumed in this standard account of science in public
policy to be distinct from values, which are seen as the primary medium of
exchange in the political realm.
Values [distinct from the scientific community’s “internal” normative
commitments] are thought to play no significant role in the creation of
scientific facts. Realists believe
that productive discussion of norms and values stops at the point where public
choices come to depend chiefly on experts’ objective assessments of the
facts. By extension it is the duty
of expert policy advisers to bring facts to bear on the processes of political
evaluation and judgment, and so to keep public actions from falling prey to
passion and irrationality.
(Jasanoff 1997, p. 229)
From such a ‘realist’ perspective, one would expect the policy-relevance
of scientific consensus to correlate (more or less) with the “internal”
normativity of the scientific consensus: the closer the scientific-consensus
process conforms to the norms of scientific practice, the more ‘objective’ will
be the scientific consensus reached and, hence, the greater will be the
relevance of that consensus in any ‘rational’ public policy-making
process.
Our study of this case has shown mixed results for the ‘realist’
position. In the case of North
Carolina, the failures of the norms of disinterestedness, communalism and
organized skepticism, due to the influence of politics on science, did
significantly compromise the ability of scientific consensus to inform public
policy, in accord with the ‘realist’ view. Concerning the river closure
policy in Maryland, however, the initial short-term policy-relevance of
scientific consensus came at the cost of the organized skepticism of the
scientific community, although in the longer term further research did help
refine public policy. And comparing
the cases of Maryland and North Carolina, we find an anti-correlation between
the universality of the scientific consensus and the policy-relevance of that
consensus, in contrast to the ‘realist’ view. Thus, on the basis of this study, one
would agree that ‘realism’ is at least somewhat naïve in its understanding of
the relationship between science and policy and for sure inadequate to
comprehend the science-policy interaction in the present case.
4.2.
Science irrelevant to policy—radical ‘skepticism’
Given
the mismatch between science and politics as evident in this particular case,
one might conclude that not only was science in this case actually irrelevant to
policy but further that science generally is, in fact, and ought to be,
irrelevant to policy. Such is the
radically ‘skeptical’ view of Collingridge and Reeve (1986), who reject the
rationalism of the ‘realist’ view.
They make two central claims: first, that “relevance to policy, by
itself, is sufficient to completely destroy the delicate mechanisms by which
scientists normally ensure that their work leads to agreement” and, second, that
“this failure of science to operate as smoothly as mythology would have it does
not in effect matter for policymaking” (Collingridge and Reeve 1986, pp.
ix-x). In their view, science can
relate to the policy making process in either of two ways, but in both ways
science not only fails to live up to its best ideals but also has little real
effect on policy decisions. In the
first, science is called in to settle a policy debate by reaching consensus on
new evidence, but in attempting to do so becomes dysfunctionally mired in
hyper-critical technical disputes that provide no guidance for policy makers who
end up reaching decisions largely independently of the scientific debate. In the second, an existing
‘under-critical’ scientific consensus ends up merely confirming a prior policy
consensus by failing to give equal weight to alternative hypotheses. They thus draw the conclusion that “the
corrigibility of all scientific claims means that policy ought to be insensitive
to any conjecture of science” (p. 29).
There
is no doubt that the case of North Carolina provides ample evidence supporting
their first claim that political relevance can indeed distort and disrupt the
normal functioning of science to the extent of undercutting the ability of
scientific consensus to objectively inform public policy. Is the present case also evidence in
favor of their second claim that science is and ought to be irrelevant to
policy? That depends on the
inevitability of such irrelevance of science to policy. We will test their claim against the
case of Maryland.
Regarding
river closures, the early policy followed by state officials was motivated by a
medically warranted concern over the potential human health risk from Pfiesteria toxicity and was keyed to an
assumed linkage between toxic Pfiesteria activity and observable fish
lesions. This latter assumption was
supported by the early TAC consensus, which, despite noted uncertainties, gave
scientific objectivity to a policy that was publicly controversial because of
its impact on private economic and recreational activities. During the next two years, however, the
Pfiesteria-lesion linkage came under
considerable scrutiny within the scientific community, resulting in a fair
amount of dispute among researchers concerning the etiology of the fish lesions
that had been previously attributed to Pfiesteria as the primary cause. The TAC reconvened in late 1998 and in
light of the new research significantly qualified its earlier conclusions: “This
uncertainty does mean…that the prevalence of fish lesions alone should not be
considered a reliable indicator of toxic Pfiesteria outbreaks” (Center for
Environmental Science 1999).
Consequently, the state’s river closure policy came under both scientific
and public criticism as the results of these further fish pathology studies were
released. State officials,
acknowledging that “since many factors may cause lesions in fish” subsequently
revised its river closure policy by abandoning lesions as the primary and
exclusive environmental indicator of recent toxic Pfiesteria activity in favor of multiple
indicators, including fish behavior and visual identification of the organism in
addition to lesions (Department of Natural Resources 1999). For sure, the Pfiesteria-lesion linkage question has
generated hyper-critical technical disputes among scientists as in the first
scenario depicted above.
Instead
of this dispute hindering public policy or rendering science irrelevant to
policy, however, it rather prompted a revision of public policy to bring it more
in line with the limits of present knowledge; and this was so because the
original policy was explicitly left open to “continuous evaluation” and revision
“in light of improvements to our knowledge” (Maryland Department of Natural
Rescources 1997). Now, the
science in dispute was not the ultimate basis of the river closure policy, which
was motivated primarily by political risk assessment and would likely have been
made by state officials on the basis of the available medical evidence and
standard public health precaution irrespective of the question of lesion
etiology, the latter of which effectively mattered regarding only under what
circumstances the policy should be invoked. So, given that this policy was justified
based on public values more than scientific evidence to begin with, it is
perhaps not the best issue on which to test the Collingridge and Reeve
claim.
Looking
on the face of things regarding water quality and nutrient control, one might be
led to conclude that, in fact, science was irrelevant to policy in that
area. The policy-making process on
water quality, bolstered politically by an early (uncertainty-laden) scientific
consensus in favor of the Pfiesteria-lesion linkage, moved quickly
out ahead of the peer-review process on lesion etiology in the direction of
stronger nutrient control measures, in favor of which there was already a
substantial consensus among environmental advocates, scientists, and
managers. A second expert
scientific committee, practically restricted by the mandate of the policy-making
process, drew up a scientific consensus that effectively validated the prior
policy consensus on nutrient control and consciously framed the Pfiesteria-nutrients issue within this
prior consensus. The policy-making
process then ran with this conclusion on its way toward new regulations on
agricultural land use. Moreover,
the public policy finally enacted doesn’t appear to be based upon Pfiesteria itself in that it reflects
neither the extent of the original Pfiesteria-related event (a relatively
minor fish kill that affected a small, remote Bay tributary), nor the science of
Pfiesteria (which suggested that Pfiesteria thrives in only nutrient rich
waters), nor even the actual public health risk associated with Pfiesteria (which is effectively limited
to direct contact with those specific waters where recent toxic Pfiesteria activity is suspected). Instead, the WQIA mandates nutrient
management planning on (nearly) every farm in every tributary basin regardless
of the susceptibility to toxic Pfiesteria blooms (which appear
restricted to shallow, slow moving brackish waters), likelihood of fish kills
(which chiefly affect juvenile Atlantic menhaden), relative contribution of
agricultural run-off to nutrient loads, and trends in nutrient levels (which
were much higher on the Eastern Shore tributaries) in the respective
rivers. The major features of the
policy appear rather insensitive to Pfiesteria itself, reflecting rather a
prior policy agenda on controlling non-point source nutrient pollution from
agricultural activities and a political process that must seek equity by
treating all farmers alike regardless of scientific reasons for a more
“geographically-logical” policy.
All of this would seem to fit the second scenario above, the
‘under-critical’ model.
While
on the face of it the peculiar role played by Pfiesteria in the policy-making process
in Maryland does indeed fit this scenario, it would be hasty to conclude that
science was irrelevant to the WQIA.
For the primary scientific basis for the policy was not so much Pfiesteria as the three decades of prior
research on nutrient pollution, estuarine eutrophication and algal blooms. Now, the nutrient reduction agenda
surely did constitute a prior policy consensus, but such consensus was motivated
by that prior research independently of Pfiesteria; and that prior research
consensus is hardly what one would consider ‘under-critical.’ Rather than the ‘under-critical’ model
of Collingridge and Reeve, the role of Pfiesteria better fits what Dr. Michael
Orbach, Director of the Duke Marine Biology Laboratory, refers to as the
‘lever.’ Sometimes in environmental
advocacy, one issue is used as a ‘lever’ for another issue. In such cases “the lever can be
scientifically weak, but it might have other factors that make it usable for
advocacy purposes.”[33] The human health, economic and cultural
impacts of Pfiesteria were used by
advocates for stronger nutrient control to leverage mandatory nutrient
management planning even though the scientific linkages of Pfiesteria to lesions and fish health
were significantly uncertain. So,
science entered into the water quality policy-making process in Maryland at two
levels—Pfiesteria as the short-term
lever for a policy warranted by long-term nutrient research. There was science, then, that was
relevant to the water quality policy-making process in an epistemologically
defensible way that did not necessarily betray the ideals of science; and
without that long-term research program, such water quality policy would likely
not have been on the table in the first place.
From
the perspective of Collingridge and Reeve, one would say that the case of the
river closure policy fits the ‘over-critical’ scenario, while the case of the
nutrient management regulations fits the ‘under-critical’ scenario.