Bending Science
Outcome-driven research
Outcome-driven research
Restoring Science: Forcing Bent Science Out Into the Open
"In bending science, the casualties are not limited to the powerless members of a diffuse public and the fragile democratic processes highlighted in the political science literature and editorial pages. Some of the largest casualties of bent science include the bedrock institutions of science itself. Scientists on the margins of professional respectability thrive as their well-supported, outcome-driven research dominates some field of policy-relevant science, while other scientists honestly practicing their craft find themselves sorted into camps by special interests who either praise their work and support it or attack it and attempt to stifle it. The editors of our best international scientific and medical journals are chagrined by their inability to weed out unreliable research emerging from a funding regime that is increasingly driven by the expectation of future economic gain. Having implemented strict conflict of interest policies, they find that they lack sufficient resources and influence to ensure meaningful compliance. Even our great research universities are threatened, as evidence mounts that they frequently operate more like profit-driven corporations than institutions dedicated first and foremost to the pursuit and transmission of knowledge. The public's faith in science, while never easy to measure, may finally be eroding under the steady flow of reports of manipulated and distorted research." (Source: Bending Science: How special interests corrupt public health research. "Chapter 10. Restoring Science: Forcing Bent Science Out into the Open" p. 229-230. Thomas O. McGarity and Wendy E. Wagner. Harvard University Press. Cambridge, Massachusetts, USA. 2008. ISBN: 978-0-674-04714-3. [Book Review: The Journal of Clinical Investigation or pdf])
"In bending science, the casualties are not limited to the powerless members of a diffuse public and the fragile democratic processes highlighted in the political science literature and editorial pages. Some of the largest casualties of bent science include the bedrock institutions of science itself. Scientists on the margins of professional respectability thrive as their well-supported, outcome-driven research dominates some field of policy-relevant science, while other scientists honestly practicing their craft find themselves sorted into camps by special interests who either praise their work and support it or attack it and attempt to stifle it. The editors of our best international scientific and medical journals are chagrined by their inability to weed out unreliable research emerging from a funding regime that is increasingly driven by the expectation of future economic gain. Having implemented strict conflict of interest policies, they find that they lack sufficient resources and influence to ensure meaningful compliance. Even our great research universities are threatened, as evidence mounts that they frequently operate more like profit-driven corporations than institutions dedicated first and foremost to the pursuit and transmission of knowledge. The public's faith in science, while never easy to measure, may finally be eroding under the steady flow of reports of manipulated and distorted research." (Source: Bending Science: How special interests corrupt public health research. "Chapter 10. Restoring Science: Forcing Bent Science Out into the Open" p. 229-230. Thomas O. McGarity and Wendy E. Wagner. Harvard University Press. Cambridge, Massachusetts, USA. 2008. ISBN: 978-0-674-04714-3. [Book Review: The Journal of Clinical Investigation or pdf])
Bending Policy-relevant health and environmental research
"As Thomas McGarity and Wendy Wagner demonstrate quite vividly, advocates frequently distort policy-relevant health and environmental research – that is, bend science – in order to serve economic or ideological interests. This elaborate charade plays out in the halls of Congress, the regulatory arena, the courtroom, and the media. To some readers, including many scientists with no direct professional familiarity with the world of policy, the well-honed message of this book may be genuinely shocking. To scientists whose work may have intersected occasionally with the policy realm, the message will not surprise, but the pervasiveness of the phenomenon is likely to astonish, and worry." (Source: "Bent science" by Kenneth E. Warner. Issues in Science and Technology. Spring 2009. p. 90-92. Review of: Bending Science: How Special Interests Corrupt Public Health Research by Thomas O. McGarity and Wendy E. Wagner. Cambridge, MA: Harvard University Press. 2008, 400 pp.)
"As Thomas McGarity and Wendy Wagner demonstrate quite vividly, advocates frequently distort policy-relevant health and environmental research – that is, bend science – in order to serve economic or ideological interests. This elaborate charade plays out in the halls of Congress, the regulatory arena, the courtroom, and the media. To some readers, including many scientists with no direct professional familiarity with the world of policy, the well-honed message of this book may be genuinely shocking. To scientists whose work may have intersected occasionally with the policy realm, the message will not surprise, but the pervasiveness of the phenomenon is likely to astonish, and worry." (Source: "Bent science" by Kenneth E. Warner. Issues in Science and Technology. Spring 2009. p. 90-92. Review of: Bending Science: How Special Interests Corrupt Public Health Research by Thomas O. McGarity and Wendy E. Wagner. Cambridge, MA: Harvard University Press. 2008, 400 pp.)
Bending Science to Hide Unwelcome, Inconvenient Results
The excerpts included below present a real-world example of bending science to hide unwelcome, inconvenient scientific results, conclusions and truth. The sources of this example include: 1) Bending Science: How special interests corrupt public health research and 2) Scientific American, "Doubt is Their Product" Consequently, while reading these excerpts, mentally compare NCFM Snurra/eSnurra Groups' "bending" the reporting of the risk of perinatal death which results from changing LMPD-based EDD by >=14 days with Merk Corporation hiding a "five-fold increase in the risk of heart attack" for users of the painkiller Vioxx relative to users of naproxen [sold under the brand name Aleve].
The excerpts included below present a real-world example of bending science to hide unwelcome, inconvenient scientific results, conclusions and truth. The sources of this example include: 1) Bending Science: How special interests corrupt public health research and 2) Scientific American, "Doubt is Their Product" Consequently, while reading these excerpts, mentally compare NCFM Snurra/eSnurra Groups' "bending" the reporting of the risk of perinatal death which results from changing LMPD-based EDD by >=14 days with Merk Corporation hiding a "five-fold increase in the risk of heart attack" for users of the painkiller Vioxx relative to users of naproxen [sold under the brand name Aleve].
- Bending Science "Chapter 4. Shaping Science: The Art of Creating Research to Fit One's Needs"
"A modern example of data interpretation to downplay risks is the interpretation Merck Corporation scientists provided for the results of a company-sponsored postmarketing clinical trial in which the statistical analysis showed that subjects taking Vioxx [rofecoxib] for nine months had five times the risk of heart attack as those taking naproxen [sold under the brand name Aleve], another commonly used painkiller. 77 Since the tests were not run against a control population of persons not taking any painkiller, scientists working for the manufacturer could interpret these results in one of two ways. Either taking Vioxx increased the risk of heart attacks by around 400 percent, or taking naproxen reduced the risk of heart attack by about 80 percent. Although the latter interpretation seemed highly implausible, the company’s scientists adopted it, and Vioxx remained on the market for another five years until another post marketing trial comparing Vioxx with a placebo showed that Vioxx caused a two-fold increase in the risk of heart attack. According to one FDA scientist, Vioxx caused between 80,000 and 139,000 heart attacks during those five years." (Source: Bending Science: How special interests corrupt public health research. "Chapter 4. Shaping Science: The Art of Creating Research to Fit One's Needs" p. 76. Thomas O. McGarity and Wendy E. Wagner. Harvard University Press. Cambridge, Massachusetts, USA. 2008. ISBN: 978-0-674-04714-3. [Book Review: The Journal of Clinical Investigation or pdf]) - Doubt is Their Product
"Although the Wall Street Journal has reported that certain documents suggest that Merck executives were aware of the increased risk of heart attacks, it is hard to imagine that the company’s scientists were deliberately promoting a drug they knew was unsafe [Vioxx]. At the same time, it is hard to imagine they honestly thought naproxen reduced the risk of heart attack by 80 percent. If they did, they should have urged the government to pour it straight into the water supply. It seems more likely that their allegiances were so tightly linked with the products they worked on, as well as the financial health of their employers, that their judgment became fatally impaired. And the FDA? That agency has neither the legal authority nor the resources to effectively identify the adverse outcomes caused by drugs already on the market." (Source: "Doubt Is Their Product: Industry groups are fighting government regulation by fomenting scientific uncertainty" David Michaels. Scientific American June 2005. p. 96-101. PDF) - "Of all the tools that affected parties employ to distort science, the tactic of “hiding” science is the most straightforward. With the possible exception of shaping science through overt scientific fraud, hiding science presents the most dramatic affront to both science and informed public policy-making. The numerous examples of suppression that follow suggest that it may also be the most frequently employed strategy for bending science. Given its very nature, it seems likely that many more accounts remain suppressed than are uncovered. From the very beginnings of modern science to the present day, sponsors of scientific research have attempted to hide results that have adverse implications for their economic, political or ideological positions, and we have not identified any discernible historical trends. Since the temptation to cover up inconvenient information is universal, the incidence of hiding will depend on the availability of opportunities to hide, and the evidence suggests that the opportunities to hide science are legion." (Source: Bending Science: How special interests corrupt public health research. "Chapter 4. Hiding Science: The Art of Concealing Unwelcome Information" p. 101-102. Thomas O. McGarity and Wendy E. Wagner. Harvard University Press. Cambridge, Massachusetts, USA. 2008. ISBN: 978-0-674-04714-3. Book Review: The Journal of Clinical Investigation or pdf)
Science and Policy: Separatist View (Figure 1) vs. Bending Science View (Figure 2)
"In what we characterize as a separatist view of the world (Figure 1) scientific research is sufficiently reliable for use in legal proceedings and important public policy deliberations after it flows out of the realm of science through a pipeline in which those inhabiting the realm of science (often referred to as the “scientific community”) have the opportunity to screen the relevant scientific studies to ensure that they are produced in accordance with the norms and procedures of science. Within the pipeline, the following scenario ordinarily plays out: scientists conduct research in accordance with a predetermined research protocol; they gather the relevant data; they analyze and interpret the data; they write up the data, analysis and conclusions and submit a manuscript to a scientific journal for publication; the journal, prior to acceptance, subjects the paper to peer review; the scientists respond to the reviews by making any necessary changes; and the journal publishes the paper. If the paper has important scientific implications and attracts sufficient attention, it will undergo additional scientific scrutiny. Other scientists may publish letters in the journals critiquing or expanding upon some aspect of the paper; it may become incorporated into a “review article” summarizing the published work in a particular area of research and drawing additional conclusions; it may play a role in a report prepared by an expert body like a “blue ribbon panel” examining a scientific topic of some social relevance; and summaries of the article or the panel report may get broadly disseminated by the media to scientists, decision-makers and the public.
The separatist view does not seem to contemplate the possibility that the otherwise legitimate process of generating and refining scientific studies within the pipeline could become contaminated by outsiders with an economic or ideological stake in the ongoing legal, regulatory and policy debates. Instead, it assumes that the “scientific community” ensures, through rigorous peer review and professional oversight, that the scientific work that exits the pipeline is reliable and generally uncontaminated by biasing influences like the economic or ideological preferences of the sponsor of the research. Separatists are quick to acknowledge that once science enters the realm of policy, it becomes the plaything of advocates. And from that point on, it is Katy-bar-the-door [a warning of the approach of trouble]. As attorneys in litigation and interest groups in regulatory proceedings and policy debates seize on research that is relevant to their causes, they manipulate it however they can to achieve preferred judicial, regulatory and legislative outcomes. 3 But, separatists seem confident that these outcome-oriented strategies are readily identified and excluded within the realm of science before they have a chance to corrupt the pipeline’s output.
We believe that this separatist view is the dominant weltanschauung [a particular philosophy or view of life] of judges, regulatory policy-makers, scientists and, to the extent that it ponders such issues, the general public. In this view, bias and suppression in science are filtered out of published studies, review letters, and consensus statements used to inform legal decisions by the quality control techniques traditionally employed in the scientific process. When an advocate for a special interest is caught distorting published studies or skewing scientific consensus panels, the incident is generally considered an isolated lapse in oversight and not the result of a larger pattern of distortion.
"In what we characterize as a separatist view of the world (Figure 1) scientific research is sufficiently reliable for use in legal proceedings and important public policy deliberations after it flows out of the realm of science through a pipeline in which those inhabiting the realm of science (often referred to as the “scientific community”) have the opportunity to screen the relevant scientific studies to ensure that they are produced in accordance with the norms and procedures of science. Within the pipeline, the following scenario ordinarily plays out: scientists conduct research in accordance with a predetermined research protocol; they gather the relevant data; they analyze and interpret the data; they write up the data, analysis and conclusions and submit a manuscript to a scientific journal for publication; the journal, prior to acceptance, subjects the paper to peer review; the scientists respond to the reviews by making any necessary changes; and the journal publishes the paper. If the paper has important scientific implications and attracts sufficient attention, it will undergo additional scientific scrutiny. Other scientists may publish letters in the journals critiquing or expanding upon some aspect of the paper; it may become incorporated into a “review article” summarizing the published work in a particular area of research and drawing additional conclusions; it may play a role in a report prepared by an expert body like a “blue ribbon panel” examining a scientific topic of some social relevance; and summaries of the article or the panel report may get broadly disseminated by the media to scientists, decision-makers and the public.
The separatist view does not seem to contemplate the possibility that the otherwise legitimate process of generating and refining scientific studies within the pipeline could become contaminated by outsiders with an economic or ideological stake in the ongoing legal, regulatory and policy debates. Instead, it assumes that the “scientific community” ensures, through rigorous peer review and professional oversight, that the scientific work that exits the pipeline is reliable and generally uncontaminated by biasing influences like the economic or ideological preferences of the sponsor of the research. Separatists are quick to acknowledge that once science enters the realm of policy, it becomes the plaything of advocates. And from that point on, it is Katy-bar-the-door [a warning of the approach of trouble]. As attorneys in litigation and interest groups in regulatory proceedings and policy debates seize on research that is relevant to their causes, they manipulate it however they can to achieve preferred judicial, regulatory and legislative outcomes. 3 But, separatists seem confident that these outcome-oriented strategies are readily identified and excluded within the realm of science before they have a chance to corrupt the pipeline’s output.
We believe that this separatist view is the dominant weltanschauung [a particular philosophy or view of life] of judges, regulatory policy-makers, scientists and, to the extent that it ponders such issues, the general public. In this view, bias and suppression in science are filtered out of published studies, review letters, and consensus statements used to inform legal decisions by the quality control techniques traditionally employed in the scientific process. When an advocate for a special interest is caught distorting published studies or skewing scientific consensus panels, the incident is generally considered an isolated lapse in oversight and not the result of a larger pattern of distortion.
Scientific Pipeline: Conduct research > Interpret data > Journal peer review > Post-publish scrutiny > Expert panel opinions
(Source: Bending Science: How special interests corrupt public health research. "Chapter 1. Introduction, The Problem" p. 4-5. Thomas O. McGarity and Wendy E. Wagner. Harvard University Press. Cambridge, Massachusetts, USA. 2008. ISBN: 978-0-674-04714-3. [Book Review: The Journal of Clinical Investigation or pdf])
If it accomplishes nothing else, this book [Bending Science: How special interests corrupt public health research] should demonstrate that the separatist view is an idyllic, even Polyannaish view of the interaction between science and policy in the hotly contested areas of public health and environmental regulation and litigation. In the real world, 4 each step in the science pipeline can become seriously contaminated by the efforts of advocates who are responding to incentives and pressures from within the realm of policy. In this more realistic view, advocates infiltrate the pipeline and secretly situate themselves at various key points to ensure that the research, the critiques, the panel reports, and the overall messages match their ends. In other cases, they undermine the work of scientists who persist in conducting research that runs contrary to their goals. After the science exits the pipeline, the advocates continue their efforts but shift their focus to manipulate public perceptions about the research as it becomes available for use in the policy realm. In this alternate view, science no longer proceeds in a straight and insulated path through a closely supervised pipeline of scientific oversight; the pipeline is instead much more porous and vulnerable to a range of tricks developed by determined advocates to bend science to their own ends. Figure 2 charts out the numerous points of entry where special interests can infiltrate the pipeline and manipulate the normal processes of science, often without detection by the scientific assembly line of internal checks and review stations. Instead of a process that has managed to insulate itself from the unruly world of policy, then, sacred scientific processes are contaminated by determined advocates who understand the power of supportive science in legal and policy proceedings. Indeed, since published results and peer review statements carry the highest indicia of reliability and authority, it is on these products that advocates tend to focus most intensely.
Perhaps because they assume that the pipeline produces results that are largely insulated from contamination by advocates, the institutions of government––courts, executive branch agencies and the Congress––employ processes that not only neglect to deter efforts by advocates to bend science but might actually make it worse. Each of these legal institutions depends heavily on advocates to select the scientific information used in their decisions, and they employ surprisingly few mechanisms to screen out unscientific outcome-oriented research. Worse, the adversarial nature of the legal system actually encourages advocates to invade the realm of science––urreptitiously, because they are not especially welcome there––to commission research and manipulate scientific quality control procedures in outcome-oriented ways.
Ironically, this in turn creates even greater distance between independent researchers and bent science. Since outcome-oriented research is antithetical to science, independent scientists generally avoid both the science and the scientists who they know to be involved in producing it. This leads to an adverse selection of sorts, with the institutions in the policy realm equipped with only limited trusted expert assistance in separating the legitimate scientific research exiting the pipeline from research that has become contaminated by advocates. In many cases, scientists who leave the scientific realm to make their expertise available to institutions in the policy realm are in fact employed by advocates for the purpose of advocacy, not for the purpose of helping governmental institutions arrive at the scientific truth. Worst of all, some independent scientists whose work informs policy may find themselves on the receiving end of harassing attacks from other scientists who, unbeknownst to them, have been hired by advocates to deconstruct their work.
The reluctance of independent scientists to involve themselves in the scientific review of and policy debates over outcome-oriented research plays into the hands of the advocates because it allows them to maintain the separatist illusion, knowing full-well that it does not reflect reality. Although the advocates did not by any means originate the separatist worldview, they carefully nurture it in the realm of policy because it allows them to take the position that decisions in the policy realm should be based on “sound science,” rather than on precautionary policies, while at the same time retaining a powerful influence over the “science” that drives those decisions. The final result is a train wreck of sorts with regard to a broad range of public health and environmental issues as inaccurate separatist idealizations of science collide with the reality of “science-based advocacy.” This book documents the collision and explains the forces that have produced it. The final three chapters offer suggestions for cleaning up the resulting mess and avoiding similar collisions in the future.
How Advocates Bend Science
To understand why advocates’ influence on scientific research can be both illegitimate and easily accomplished, we must begin with the scientific ideal to which researchers in the realm of science aspire. Science is knowledge that results from testing a hypothesis with an open mind using methods that scientists have accepted as valid and generally capable of replication. 5 While philosophers and sociologists of science may debate some of the precise qualities that define science, they all agree that research conducted with a predetermined outcome is not science. 6 Indeed, the productivity of the scientific enterprise depends on the commitment of scientists to perform and critique research in a disinterested fashion, meaning at the very least that they do not manipulate their methods and analyses to produce predetermined outcomes. 7 Scientific studies must be replicable, and many are in fact replicated. But when scientists can build on the objective efforts of trusted peers without pausing to replicate that work, they can advance the scientific enterprise without wasting precious research resources. Indeed many individual scientists are so wedded to this norm of open-mindedness that they decline to participate in any legal or political activity, at least insofar as their research is concerned, to ensure that their findings are not tainted by even the appearance of outcome-oriented bias. 8 Disclosures of conflicts of interest have also become standard in science, with some of the most prestigious journals denying publication rights and status as peer reviewers to scientists who appear significantly compromised by sponsor bias. 9
The scientific ideal is, of course, just that, and deviations are not uncommon, even in independently conducted basic research. Scientists face pressures to bend research in outcome-oriented directions quite apart from incentives that emanate from the realm of policy. Scientific careers in academia and industry depend upon successful completion of research and, in the case of academia, frequent publication of research output. Research that yields unexpected outcomes or “breakthroughs” can generate great rewards, both reputational and financial, for the scientists who produce the research. Scientists are not above yielding to these pressures by surreptitiously departing from scientific norms. 10 Indeed, reports of overt scientific fraud have stimulated activity in the legal system to deal with such issues in the context of publicly supported research. 11 Even apart from these rare instances of overt fraud, well-intentioned scientists are human beings with their own biases that may be reflected in their work product. 12 Many scientists have formed very strong opinions about how the universe functions in the particular micro-segment of the universe to which they devote their research efforts, and they may consciously or unconsciously avoid or denigrate research outcomes that vary from those preconceived opinions. 13 A scientist who has demonstrated a commitment to a particular view of the world is, of course, a prime target for advocates who are always on the lookout for experts whose views line up with their economic or ideological interests and who might be willing to engage in research, sit on government advisory committees, and serve as expert witnesses.
The bias in scientific research that results when the scientist works backwards from a pre-ordained result is outside the realm of legitimate science because it lacks a fundamental characteristic of science – the open-minded pursuit of truth. When a scientist engages in such outcome-oriented research at the behest of an advocate who has intervened into an otherwise legitimate scientific process, we may properly characterize the output as “bent research” – rather than “independent” research -- to reflect the fact that the scientist has bent it in the direction of an outcome that advances the advocate’s economic or ideological interests. The mere fact that research has received support from an entity that is an advocate does not, standing alone, render it “bent,” so long as the scientist pursues it independently in accordance with the norms and procedures of science. To be relegated to the category of “bent research,” it must result from a process in which an advocate has exerted direct or strong indirect control with an eye toward producing a result favorable to its interests. 14
Bent science can also occur at later stages of the research pipeline. Scientists regularly interpret and critique the research of other scientists, write review articles and books incorporating many individual studies into a broader body of research, and sit on advisory committees that summarize research in reports designed to be comprehensible to lay decisionmakers and the public. When advocates intervene into this process by commissioning books and review articles, assembling “blue ribbon” panels, or nominating scientists whose views comport with their interests to advisory committees, they are bending science in an indirect and perhaps less influential way than when they bend the research itself, but the output is nevertheless bent science.
Advocates and the scientists who work with them employ one or more of a number of strategies to bend science in outcome-oriented ways. (Figure 3). At the most general level, they can bend research, and they can bend perceptions about research. When advocates attempt to bend research in the realm of science, their goal is to ensure that the research that exits the pipeline into the policy realm will be useful (or at least not harmful) to their efforts to influence common law courts, regulatory agencies, legislatures and other governmental bodies that have the power to affect their economic or ideological interests. More specifically, an advocate can shape research by commissioning studies that work backwards from the desired outcome through the processes of data interpretation, data analysis, data gathering, and study design. The outcome-determined results are then available for use in legal decision-making and public policy deliberations. If the commissioned research fails to produce the desired results, the advocate who commissioned it can then hide it by either halting it if its unwelcome aspects become apparent soon enough, or by keeping it out of the pipeline altogether." (Source: Bending Science: How special interests corrupt public health research. "Chapter 1. Introduction" p. 3-11. Thomas O. McGarity and Wendy E. Wagner. Harvard University Press. Cambridge, Massachusetts, USA. ISBN: 978-0-674-04714-3. Book Review: The Journal of Clinical Investigation or PDF)
Perhaps because they assume that the pipeline produces results that are largely insulated from contamination by advocates, the institutions of government––courts, executive branch agencies and the Congress––employ processes that not only neglect to deter efforts by advocates to bend science but might actually make it worse. Each of these legal institutions depends heavily on advocates to select the scientific information used in their decisions, and they employ surprisingly few mechanisms to screen out unscientific outcome-oriented research. Worse, the adversarial nature of the legal system actually encourages advocates to invade the realm of science––urreptitiously, because they are not especially welcome there––to commission research and manipulate scientific quality control procedures in outcome-oriented ways.
Ironically, this in turn creates even greater distance between independent researchers and bent science. Since outcome-oriented research is antithetical to science, independent scientists generally avoid both the science and the scientists who they know to be involved in producing it. This leads to an adverse selection of sorts, with the institutions in the policy realm equipped with only limited trusted expert assistance in separating the legitimate scientific research exiting the pipeline from research that has become contaminated by advocates. In many cases, scientists who leave the scientific realm to make their expertise available to institutions in the policy realm are in fact employed by advocates for the purpose of advocacy, not for the purpose of helping governmental institutions arrive at the scientific truth. Worst of all, some independent scientists whose work informs policy may find themselves on the receiving end of harassing attacks from other scientists who, unbeknownst to them, have been hired by advocates to deconstruct their work.
The reluctance of independent scientists to involve themselves in the scientific review of and policy debates over outcome-oriented research plays into the hands of the advocates because it allows them to maintain the separatist illusion, knowing full-well that it does not reflect reality. Although the advocates did not by any means originate the separatist worldview, they carefully nurture it in the realm of policy because it allows them to take the position that decisions in the policy realm should be based on “sound science,” rather than on precautionary policies, while at the same time retaining a powerful influence over the “science” that drives those decisions. The final result is a train wreck of sorts with regard to a broad range of public health and environmental issues as inaccurate separatist idealizations of science collide with the reality of “science-based advocacy.” This book documents the collision and explains the forces that have produced it. The final three chapters offer suggestions for cleaning up the resulting mess and avoiding similar collisions in the future.
How Advocates Bend Science
To understand why advocates’ influence on scientific research can be both illegitimate and easily accomplished, we must begin with the scientific ideal to which researchers in the realm of science aspire. Science is knowledge that results from testing a hypothesis with an open mind using methods that scientists have accepted as valid and generally capable of replication. 5 While philosophers and sociologists of science may debate some of the precise qualities that define science, they all agree that research conducted with a predetermined outcome is not science. 6 Indeed, the productivity of the scientific enterprise depends on the commitment of scientists to perform and critique research in a disinterested fashion, meaning at the very least that they do not manipulate their methods and analyses to produce predetermined outcomes. 7 Scientific studies must be replicable, and many are in fact replicated. But when scientists can build on the objective efforts of trusted peers without pausing to replicate that work, they can advance the scientific enterprise without wasting precious research resources. Indeed many individual scientists are so wedded to this norm of open-mindedness that they decline to participate in any legal or political activity, at least insofar as their research is concerned, to ensure that their findings are not tainted by even the appearance of outcome-oriented bias. 8 Disclosures of conflicts of interest have also become standard in science, with some of the most prestigious journals denying publication rights and status as peer reviewers to scientists who appear significantly compromised by sponsor bias. 9
The scientific ideal is, of course, just that, and deviations are not uncommon, even in independently conducted basic research. Scientists face pressures to bend research in outcome-oriented directions quite apart from incentives that emanate from the realm of policy. Scientific careers in academia and industry depend upon successful completion of research and, in the case of academia, frequent publication of research output. Research that yields unexpected outcomes or “breakthroughs” can generate great rewards, both reputational and financial, for the scientists who produce the research. Scientists are not above yielding to these pressures by surreptitiously departing from scientific norms. 10 Indeed, reports of overt scientific fraud have stimulated activity in the legal system to deal with such issues in the context of publicly supported research. 11 Even apart from these rare instances of overt fraud, well-intentioned scientists are human beings with their own biases that may be reflected in their work product. 12 Many scientists have formed very strong opinions about how the universe functions in the particular micro-segment of the universe to which they devote their research efforts, and they may consciously or unconsciously avoid or denigrate research outcomes that vary from those preconceived opinions. 13 A scientist who has demonstrated a commitment to a particular view of the world is, of course, a prime target for advocates who are always on the lookout for experts whose views line up with their economic or ideological interests and who might be willing to engage in research, sit on government advisory committees, and serve as expert witnesses.
The bias in scientific research that results when the scientist works backwards from a pre-ordained result is outside the realm of legitimate science because it lacks a fundamental characteristic of science – the open-minded pursuit of truth. When a scientist engages in such outcome-oriented research at the behest of an advocate who has intervened into an otherwise legitimate scientific process, we may properly characterize the output as “bent research” – rather than “independent” research -- to reflect the fact that the scientist has bent it in the direction of an outcome that advances the advocate’s economic or ideological interests. The mere fact that research has received support from an entity that is an advocate does not, standing alone, render it “bent,” so long as the scientist pursues it independently in accordance with the norms and procedures of science. To be relegated to the category of “bent research,” it must result from a process in which an advocate has exerted direct or strong indirect control with an eye toward producing a result favorable to its interests. 14
Bent science can also occur at later stages of the research pipeline. Scientists regularly interpret and critique the research of other scientists, write review articles and books incorporating many individual studies into a broader body of research, and sit on advisory committees that summarize research in reports designed to be comprehensible to lay decisionmakers and the public. When advocates intervene into this process by commissioning books and review articles, assembling “blue ribbon” panels, or nominating scientists whose views comport with their interests to advisory committees, they are bending science in an indirect and perhaps less influential way than when they bend the research itself, but the output is nevertheless bent science.
Advocates and the scientists who work with them employ one or more of a number of strategies to bend science in outcome-oriented ways. (Figure 3). At the most general level, they can bend research, and they can bend perceptions about research. When advocates attempt to bend research in the realm of science, their goal is to ensure that the research that exits the pipeline into the policy realm will be useful (or at least not harmful) to their efforts to influence common law courts, regulatory agencies, legislatures and other governmental bodies that have the power to affect their economic or ideological interests. More specifically, an advocate can shape research by commissioning studies that work backwards from the desired outcome through the processes of data interpretation, data analysis, data gathering, and study design. The outcome-determined results are then available for use in legal decision-making and public policy deliberations. If the commissioned research fails to produce the desired results, the advocate who commissioned it can then hide it by either halting it if its unwelcome aspects become apparent soon enough, or by keeping it out of the pipeline altogether." (Source: Bending Science: How special interests corrupt public health research. "Chapter 1. Introduction" p. 3-11. Thomas O. McGarity and Wendy E. Wagner. Harvard University Press. Cambridge, Massachusetts, USA. ISBN: 978-0-674-04714-3. Book Review: The Journal of Clinical Investigation or PDF)
Figure 3: The Tools for Bending Science
(Source: Bending Science: How special interests corrupt public health research. "Chapter 1. Introduction, The Problem" p. 10. Thomas O. McGarity and Wendy E. Wagner. Harvard University Press. Cambridge, Massachusetts, USA. 2008. ISBN: 978-0-674-04714-3. Google Scholar: Book Review: The Journal of Clinical Investigation or pdf)
The captions of the shelf items in Figure 3 above are included below for legibility.
Spinning
|
(Source: Bending Science: How special interests corrupt public health research. "Chapter 1. Introduction, The Problem" p. 10. Thomas O. McGarity and Wendy E. Wagner. Harvard University Press. Cambridge, Massachusetts, USA. 2008. ISBN: 978-0-674-04714-3. Google Scholar: Book Review: The Journal of Clinical Investigation or pdf)