Commentary: Science, Policy, and The Transparency of Values
Commentary: Science, Policy, and The Transparency of Values
Commentary: Science, Policy, and The Transparency of Values
to enter these waters because public policies that evidence of toxicity in animals is relevant be based on this form of evidence, because it
should typically be informed by the best avail- to humans. The authors argue that the alterna- may be appropriate to use different standards
able scientific evidence (Pielke 2007; Resnik tive assumption (i.e., that effects in animals do of evidence to protect the public from risks. A
2009). In a commentary that cites Dietrich not provide evidence for effects in humans) continuum of different kinds and amounts of
etal. (2013) approvingly, Lehman-McKeeman would be unworkable (Bergman etal. 2013). evidence could be demanded for policy deci-
and Kaminski (2013) argued that the Society But this conclusion incorporates value judg- sions (Ashford 1988). Very high standards of
of Toxicology (SOT) must avoid playing it ments concerning the standards of evidence evidence are typically expected in order to infer
safe. In other words, they call for the mem- that are appropriate for regulating chemicals. causal relationships or to approve the market-
bers of the SOT to inform policy makers about Insisting that chemicals should be regulated ing of new drugs. In other social contexts, such
issues on which they have expertise. However, only in response to evidence from human as tort law and chemical regulation, weaker
if toxicologists and other scientists are to help studies would help to prevent false positive standards of evidence are sometimes accept-
inform policy, they face the question of how to conclusions about chemical toxicity, but it able to protect the public (Cranor 2008). To
do this without losing their objectivity or the would also prevent society from taking effec- demand the very highest standards of evidence
publics trust. tive action to minimize the risks of chemicals for chemical regulationincluding, for exam-
before they produce measurable adverse effects ple, human evidence, accompanying animal
Discussion in humans. Moreover, insisting on human data, mechanistic evidence, and clear exposure
In the past, scientists and philosophers have studies would result in failure to identify some datawould take very long periods of time
argued that the best way to maintain sciences human health risks because the diseases are and leave the publics health at risk. Thus, the
objectivity and the publics trust is to draw a rare, or the induction and latency periods are demand that regulators rely on the same stan-
sharp line between science and human val- long, or the effects are subtle (Cranor 2011). dards of evidence for toxicity as the scientific
ues or policy (Longino 1990). However, it is Similarly, Gore etal. (2013) argued that community uses in other contexts is itself a
not possible to maintain this distinction, both [t]he assumption of no threshold has been value-laden proposal.
because values are crucial for assessing what widely used, for many years, in the regulation The value-laden assumptions about stan-
counts as sufficient evidence and because ethi- of genotoxic carcinogens, often based upon dards of evidence in this dispute over endo-
cal, political, economic, cultural, and religious invitro data. We believe extending this prec- crine disruption are similar to broader social
factors unavoidably affect scientific judgment edent to EDCs is supported by the science. disputes over the precautionary principle
(Douglas 2009; Elliott 2011; Longino 1990; But the claim that the no-threshold hypothe (Kriebel etal. 2001; Martuzzi 2007; Miller and
Resnik 2007, 2009). Insisting that science is sis is supported by the science depends on Conko 2001; Sunstein 2005). Indeed, the title
value-free, when the arguments and evidence implicit assumptions about how much scien- of the editorial by Dietrich etal. (2013) begins
show that this is an unrealistic goal, perpetu- tific evidence is needed to justify formulat- with the claim, Scientifically unfounded pre-
ates a misunderstanding that interferes with the ing policy on this basis. And the question of caution drives European Commissions recom-
publics understanding of the scientific process how much evidence is needed should depend mendations on EDC regulation. Some critics
and may, paradoxically, undermine the pub- in part on value judgments about the relative of the precautionary principle, such as Dietrich
lics trust in science. We suggest that society is benefits and harms to society of assuming (or and his coauthors, argue that precaution runs
likely to be better served when scientists strive not assuming) a threshold when performing counter to scientific principles (Miller and
to be as transparent as possible about the ways risk assessments of EDCs. In this case, past Conko 2001). But decisions about how much
that interests and values may influence their toxicological experience may support the evidence to demand before taking regulatory
judgment and reasoning, while still striving for threshold hypothesis, whereas other lines of actions necessarily incorporate both scientific
objectivity. Transparency can promote public evidence (such as the proposed molecular judgments and value judgments. Because the
trust by helping laypeople understand how mechanisms by which EDCs could disrupt scientific conventions for inferring evidence
both empirical evidence and value assumptions development and generate irreversible effects of harm in some fields might require placing
enter into scientific decision making and policy on endocrine-sensitive organs) support the the public at risk for extended periods of time
formation. As the National Research Council no-threshold hypothesis. Thus, the dispute before the evidence could be accumulated, pre-
(NRC) report Understanding Risk (NRC 1996) between Gore etal. (2013) and Dietrich etal. cautionary decisions to engage in particular
emphasized, it is usually unrealistic to keep the (2013) regarding the adoption of thresholds forms of regulation may sometimes be appro-
process of risk characterization purely value- for EDCs could be clarified if the participants priate in response to more limited evidence
free. Instead, the report called for incorporat- were more forthcoming about their assump- (Cranor 2011; Martuzzi 2007).
ing broad-based deliberation about the values tions regarding the level and kind of evidence Financial, personal, and cultural influ-
that inform risk assessments in order to pro- needed to justify adopting or rejecting the ences. A second reason to avoid trying to
vide a context for the scientific analyses that are threshold hypothesis. maintain a sharp distinction between science
part of the assessment process. In their editorial, Lehman-McKeeman and values in the policy context is that per-
Standards of evidence. The first reason and Kaminski (2013) call for the members of sonal, ethical, political, and cultural values
it is problematic to draw a sharp separation the SOT to be strong advocates for applying unavoidably influence scientific reasoning.
between science and values is that values are the best science to policy issues and to craft This point is illustrated by recent concep-
necessary to decide what standards of evidence regulatory policies that are based on sound tual and empirical literature on the ways
to demand when informing policy decisions science. Even this seemingly innocuous advice that financial relationships can affect scien
(Ashford 1988; Cranor 1993; Douglas 2009; to promote decisions based on good science tific judgment and reasoning (Dana and
Elliott 2011). Several comments from the hides significant value judgments about the Loewenstein 2003; Elliott 2008; Resnik and
recent dispute about the proposed EC policy appropriate standards of evidence in policy Elliott 2013). An investigative report found
illustrate the difficulties of trying to ignore this contexts (Ashford 1988; Cranor 1993). If the that 17 of the 18 authors of the initial edito-
necessary role for values. First, as noted earlier, best science and sound science are inter- rial by Dietrich etal. (2013) had ties to regu-
the editorial that calls for science and policy on preted to mean science that meets the high- lated industries (Horel and Bienkowski 2013).
EDCs to remain unmixed (Bergman etal. est standards of scientific evidence, then it is In response to this investigation, Dietrich
2013) insists that it is reasonable to assume not clear that regulatory policy must always replied, [w]e do not believe the discussion
on the conflicts of interests will serve any- promoting transparency are also available. values may actually damage scientific objec-
body because it takes away the focus from the For example, efforts to incorporate scientists tivity and public trust, whereas a willingness
real issue (quoted by Horel and Bienkowski from a range of different stakeholder groups to bring implicit interests or values into the
2013). Bas Blaauboer, another co-author of on government advisory bodies can help to open may be the best path to promoting good
the editorial by Dietrich etal. (2013), insisted uncover and elucidate implicit value judg- science and policy.
that it was very stupid to think that his ments in science advice and promote demo-
industry involvement influenced his opinion cratic decision making (Resnik 2009). These References
(quoted by Horel and Bienkowski 2013). efforts to uncover implicit value judgments
But psychological studies have suggested that are important, given that values can influence Ashford N. 1988. Science and values in the regulatory process.
Stat Sci 3:377383.
financial interests can exert subconscious influ- subtle decisions about research questions, Babcock L, Loewenstein G, Issacharoff S. 1997. Creating
ences on human judgment even when indi- methodologies, terminology, and models convergence: debiasing biased litigants. Law Soc Inquiry
viduals are instructed about those potential (Ashford 1988; Douglas 2009; Elliott 2011; 22:401413.
Bergman , Andersson AM, Becher G, van den Berg M,
influences and motivated to remain objective Kriebel etal. 2001). Disclosures of competing BlumbergB, Bjerregaard P, etal. 2013. Science and policy
(Babcock etal. 1997; Dana and Loewenstein financial interests and nonfinancial interests on endocrine disrupters must not be mixed: a reply to a
2003; Orlowski and Wateska 1992). (such as professional or political allegiances) common sense intervention by toxicology journal editors
[Editorial]. Environ Health 12:69; doi:10.1186/1476-069X-12-69.
Importantly, people typically underestimate also provide opportunities for more transpar- Cranor CF. 1993. Regulating Toxic Substances: A Philosophy of
the strength of these influences (Dana and ent discussions of the impact of potentially Science and the Law. New York:Oxford University Press.
Loewenstein 2003; Katz etal. 2003). Some implicit and subconscious values (Resnik and Cranor CF. 2008. Toxic Torts: Science, Law, and the Possibility
of Justice. New York:Cambridge University Press.
commentators also worry that interest groups Elliott 2013). Cranor CF. 2011. Legally Poisoned: How the Law Puts Us at Risk
with deep pockets can use their financial When scientists are aware of important from Toxicants. Cambridge, MA:Harvard University Press.
and political power to skew public debate by background assumptions or values that inform Cressey D. 2013. Journal editors trade blows over toxicology.
Nature; doi:10.1038/nature.2013.13787 [Online 20September
magnifying the influence of sympathetic sci- their work, it is valuable for them to make 2013].
entists through strategic funding efforts and these considerations explicit. They can also Dana J, Loewenstein G. 2003. A social science perspective on
public relations campaigns (McGarity and make their data publicly available and strive gifts to physicians from industry. JAMA 290:252255.
Wagner 2008; Michaels 2008). to acknowledge the range of plausible inter- Dietrich DR, von Aulock S, Marquardt H, Blaauboer B,
DekantW, KehrerJ, etal. 2013. Scientifically unfounded
Financial relationships are not the only pretations of available scientific information, precaution drives European Commissions recommenda-
factors that can influence scientists. Studies the limitations of their own conclusions, the tions on EDC regulation, while defying common sense,
of risk perception have found that even prevalence of various interpretations across the well-established science and risk assessment principles
[Editorial]. Food Chem Toxicol 62:A1A4.
among professional toxicologists, men tend scientific community, and the policy options Douglas HE. 2009. Science, Policy, and the Value-Free Ideal.
to have systematically lower perceptions supported by these different interpretations. Pittsburgh, PA:University of Pittsburgh Press.
of chemical risks than women, and those This approach has much in common with EC (European Commission). 2013. Commission Recommendation
of XXXX: Defining Criteria for Endocrine Disruptors.
employed by industry have lower perceptions Ashfords seminal call for scientists to be trans- Available: http://www.environmentalhealthnews.
of risk than those in academic settings (Slovic parent about their values (Ashford 1988) and org/ehs/news/2013/pdf-links/2013.06.11%20EDC_
etal. 1997). Some of these employment with Pielkes vision of scientists as honest Recommendation%20Commission%20Draft.pdf [accessed
17March 2014].
effects may be caused or exacerbated by the brokers who open up discussions about the Elliott KC. 2008. Scientific judgment and the limits of conflict-of-
phenomenon of group polarization, whereby range of options available to decision makers interest policies. Account Res 15:129.
people exposed primarily to those who (Pielke 2007). It may even be valuable for Elliott KC. 2011. Is a Little Pollution Good for You? Incorporating
Societal Values in Environmental Research. New
share similar views ultimately adopt more scientists to reflect on how their work fits into York:Oxford University Press.
extreme positions (e.g., Sunstein 2005). Even broader social frames or narratives so that they Gore AC, Balthazart J, Bikle D, Carpenter DO, Crews D,
cultural valuesc oncerns about equality, can anticipate how their claims are likely to be Czernichow P, etal. 2013. Policy decisions on endorine
authority, individualism, and community misinterpreted or used to promote particular disruptors should be based on science across disciplines:
a response to Dietrich etal. [Editorial]. Endocrinology
can influence individuals risk perceptions political or economic agendas (McKaughan 154:39573960.
(Kahan 2010). This evidence from the social and Elliott 2013). Grandjean P, Ozonoff D. 2013. Transparency and translation of
sciences suggests that although scientists can Although scientists are rightly taught to science in a modern world [Editorial]. Environ Health 12:70;
doi:10.1186/1476-069X-12-70.
strive for objectivity, it is unrealistic to think strive for objectivity, efforts to maintain a Horel S, Bienkowski B. 2013. Special report: scientists critical
that they can provide policy advice without sharp distinction between science and policy of EU chemical policy have industry ties. Environ Health
being influenced by a variety of subconscious are likely to be counterproductive in such News. Available: http://www.environmentalhealthnews.
org/ehs/news/2013/eu-conflict [accessed 17March 2014].
factors, such as interests and values. cases as the recent dispute over EDCs. When IPCS (International Programme on Chemical Safety). 2002.
scientific evidence is disputed and major regu- Global Assessment of the State-of-the-Science of
Conclusions latory decisions are at stake, it is unrealistic to Endocrine Disruptors. Geneva:World Health Organization,
IPCS. Available: http://www.who.int/ipcs/publications/new_
Given both that values play a crucial role in think that scientists will not be influenced by issues/endocrine_disruptors/en/ [accessed 22May 2014].
choosing standards of evidence in the policy their financial, social, political, and personal Kahan D. 2010. Fixing the communications failure. Nature
context and that values have subconscious interests or values when they offer advice to 463:296297.
Katz D, Caplan A, Merz J. 2003. All gifts large and small: toward
influences on scientific judgment that are policy makers. Moreover, judgments about an understanding of the ethics of pharmaceutical industry
impossible to eliminate completely, we sug- whether EDCs exhibit thresholds or whether gift-giving. Am J Bioeth 3:3946.
gest that the best way to do policy-relevant an alleged EDC will have adverse effects in Kriebel D, Tickner J, Epstein P, Lemons J, Levins R, LoechlerEL,
research is for scientists to be as transparent humans rest not only on scientific evidence etal. 2001. The precautionary principle in environmental
science. Environ Health Perspect 109:871876.
as possible about the ways in which inter- but also on value-laden judgments about the Lehman-McKeeman L, Kaminski N. 2013. The hazards of play-
ests and values may influence their work appropriate standards of evidence. Even calls ing it safe: perspectives on how the Society of Toxicology
(Ashford 1988). The analyticdeliberative for decisions based on sound science incor- should contribute to discussions on timely issues of human
and environmental safety [Editorial]. Toxicol Sci 136:13.
approach to risk characterization described porate implicit value judgments about the Longino HE. 1990. Science as Social Knowledge: Values and
in Understanding Risk (NRC 1996) provides appropriate standards of evidence for drawing Objectivity in Scientific Inquiry. Princeton, NJ:Princeton
one model for promoting this sort of trans- policy-relevant conclusions. In cases such as University Press.
Martuzzi M. 2007. The precautionary principle: in action for
parency, but a number of other strategies for these, efforts to suppress or hide interests or public health. Occup Environ Med 64:569570.
McGarity TO, Wagner W. 2008. Bending Science: How Special NRC (National Research Council). 1996. Understanding Risk: Resnik DB. 2009. Playing Politics with Science: Balancing
Interests Corrupt Public Health Research. Cambridge, Informing Decisions in a Democratic Society. Washington, Scientific Independence and Government Oversight. New
MA:Harvard University Press. DC:National Academies Press. York:Oxford University Press.
McKaughan DJ, Elliott KC. 2013. Backtracking and the ethics of Orlowski J, Wateska L. 1992. The effects of pharmaceutical Resnik DB, Elliott KC. 2013. Taking financial relationships into
framing: lessons from voles and vasopressin. Account Res firm inticements on physician prescribing patterns. Chest account when assessing research. Account Res 20:184205.
20:206226. 102:270273. Slovic P, Malmfors T, Mertiz C, Neil N, Purchase I. 1997.
Michaels D. 2008. Doubt Is Their Product: How Industrys Pielke RA. 2007. The Honest Broker: Making Sense of Evaluating chemical risks: results of a survey of the British
Assault on Science Threatens Your Health. New Science in Policy and Politics. Cambridge, UK:Cambridge Toxicology Society. Hum Exp Toxicol 16:289304.
York:Oxford University Press. University Press. Sunstein C. 2005. Laws of Fear: Beyond the Precautionary
Miller HI, Conko G. 2001. Precaution without principle. Nat Resnik DB. 2007. The Price of Truth: How Money Affects the Principle. New York:Cambridge University Press.
Biotechnol 19:302303. Norms of Science. New York:Oxford University Press.