A common rebuttal to scientific studies is that most scientists are funded by the government, and hence are working to a government's agenda. Sometimes the claim extends to suggest that the peer-review process is biased towards the government's opinion.
This is most common for political topics, such as Evolution or AGW.
Is there any evidence that some notable research has been biased in the findings or peer review due to government influence?
From 2001 to early 2009, the US was governed by an administration which frequently got involved in climate science.
When an oil company demanded that climate scientists working on the IPCC and the United States' own National Assessment be removed:
Exxon's wish was the CEQ's command. According to an internal e-mail obtained by Rolling Stone, Connaughton's first order of business - even before his nomination was made public - was to write his White House colleagues-to-be from his law firm of Sidley & Austin. He echoed Exxon's call that Bierbaum, the acting director of the Office of Science and Technology Policy, be "dealt." In the end, each of the scientists on Exxon's hit list was replaced. "It was clear there was a strong lobby and activity against me by some in the energy industry - especially ExxonMobil," says Watson.
The Bush administration then appointed industry lobbyists to head the White House's Council on Environmental Quality. They edited scientific reports being put out by scientists working in the administration to undermine their conclusions:
On the opening page of the chapter on climate impacts, Cooney inserted a litany of language in bold intended to cast doubt on the science: "the weakest links in our knowledge . . . a lack of understanding . . . uncertainties . . . considerable uncertainty . . . perhaps even greater uncertainty . . . regarded as tentative."
And after it turned out that those modifications weren't enough:
"We tried to put some qualifiers on that chapter in the report," Cooney told him. "We'd take the text from EPA, and then we'd add a sentence like, 'We don't really know if this is really happening.' So we tried to do it, but I can see now that we made a total mess of it."
After that, Cooney was then given veto power over scientists working within the administration:
From then on, Cooney wielded a heavier pen when editing official reports on global warming. Not content to obscure science with uncertainty, he began to rewrite the science itself. Draft documents made public by the House Oversight Committee reveal that Cooney now had veto power over federal scientists, including Richard Moss, coordinator of the Climate Change Science Program Office, and even James Mahoney, the assistant commerce secretary nominally in charge of America's climate science.
The Council on Environmental Quality also controlled which administration scientists would be able to speak to the media, and edited the statements they made:
Dr. Karl, it says, "was not allowed to comment in his written testimony that 'modern climate change is dominated by human influences,' that 'we are venturing into the unknown territory with changes in climate,' or that 'it is very likely (>95 percent probability) that humans are largely responsible for many of the observed changes in climate.' " Instead of saying that global warming "is playing" a role in increased hurricane intensity, his comment became "may play" a role.
However, in the midst of all this, the scientific community made statements that contradicted the White House position. For example, in 2006 the world's largest general scientific society, the American Association for the Advancement of Science, put out a statement saying:
The scientific evidence is clear: global climate change caused by human activities is occurring now, and it is a growing threat to society. Accumulating data from across the globe reveal a wide array of effects: rapidly melting glaciers, destabilization of major ice sheets, increases in extreme weather, rising sea level, shifts in species ranges, and more. The pace of change and the evidence of harm have increased markedly over the last five years.
In 2005 the Joint Science Academies, a group of 11 national scientific societies, including the US National Academy of Sciences, stated that:
There will always be uncertainty in understanding a system as complex as the world’s climate. However there is now strong evidence that significant global warming is occurring. The evidence comes from direct measurements of rising surface air temperatures and subsurface ocean temperatures and from phenomena such as increases in average global sea levels, retreating glaciers, and changes to many physical and biological systems. It is likely that most of the warming in recent decades can be attributed to human activities (IPCC 2001). This warming has already led to changes in the Earth's climate.
In 2003 the American Geophysical Union stated that:
Human activities are increasingly altering the Earth's climate. These effects add to natural influences that have been present over Earth's history. Scientific evidence strongly indicates that natural influences cannot explain the rapid increase in global near-surface temperatures observed during the second half of the 20th century.
An article in the New York Times gives many examples of several administrations attempting to prevent the Surgeon-General from issuing statements on politically sensitive topics:
Dr. Koop, said he had been discouraged by top officials in the Reagan administration from discussing the AIDS crisis. He did so anyway.
Dr. Satcher said that the Clinton administration discouraged him from issuing a report showing that needle-exchange programs were effective in reducing disease. He released the report anyway.
Former Surgeon General Richard H. Carmona told a Congressional panel Tuesday that top Bush administration officials repeatedly tried to weaken or suppress important public health reports because of political considerations. The administration, Dr. Carmona said, would not allow him to speak or issue reports about stem cells, emergency contraception, sex education, or prison, mental and global health issues. Top officials delayed for years and tried to “water down” a landmark report on secondhand smoke, he said. Released last year, the report concluded that even brief exposure to cigarette smoke could cause immediate harm.
Dr. Carmona described being invited to testify at the government’s nine-month racketeering trial of the tobacco industry that ended in 2005. He said top administration officials discouraged him from testifying while simultaneously telling the lead government lawyer in the case that he was not competent to testify. Dr. Carmona testified anyway.
Probably the most famous example of political interference in science is, of course, Lysenkoism in Soviet Russia, described in Martin Gardner's Fads and Fallacies in the Name of Science [copied by hand, blame me for any typos]:
Seldom before in the history of modern science has a crackpot achieved the eminence, adulation and power achieved by Trofim D. Lysenko, the Soviet Union's leading authority on evolution and heredity. Not only have his opinions been pronounced dogma by the Kremlin, but his Russian opponents (whose views are held everywhere but in the USSR) in recent years have been systematically eliminated from their posts. Some have died in prison camps. Some have simply vanished. A few remain at work - but it is work in other fields of biology...
A host of experiments have been designed to test Lamarckianism. All that have been verified have proved negative. On the other hand, tens of thousands of experiments - reported in journals and carefully checked and rechecked by genticists throughout the world - have established the correctness of the gene-mutation theory beyond all reasonable doubt...
Just as Lamarckianism combines easily with an idealism in which the entire creation is fulfilling God's vast plan by constant upward striving, so also does it combine easily with political doctines which emphasize the building of a better world...
"In 1933 or thereabouts," Muller wrote, "the geneticists Chetverikoff, Ferry, and Ephroimson were all, on separate occasions, banished to Sibera, and Levitsky to a labor camp in the European Arctic... in 1936, the Communist geneticist Agol was convicted of 'Menshevik idealism' in genetics... Certain it is, however, that from 1936 on Soviet geneticists of all ranks lived a life a terror. Most of them who were not imprisoned, banished, or executed were forced to enter other lines of work.
In an expression of grim humour, Tony Judt observed in "A History of Europe Since 1945":
Stalin left his nuclear physicists alone... [He] may well have been mad but he was not stupid.
Whilst the Soviet government was able to use coercive pressure on scientists within the Soviet Union, Lysenkoism never caught on outside the reach of the Soviet bloc, and eventually collapsed even within it.
These are just a few examples - clearly politicians (on both sides of politics) do frequently attempt to interfere with science - ranging from outright oppression to limiting speech or editing reports - where the scientific results went against their political positions. However, it does seem that again and again the independent, international scientific community has been robust to government interference and influence.
In the US, the Climate Action Report, was still largely seen as in support of climate science in spite of the bits where extra uncertainty was edited in for political reasons. The Surgeons-General were still able to publish health warnings even when the government were able to oppose it. And the ultimate judge will be the independent scientific community, who continued to strongly support climate science even when the government of the day was trying to edit it out.
Are scientists biased towards the government's influence?
When a notable research has been biased due to some external influence. The is referred to as Research Bias or Experimental Bias:
Research bias, also called experimenter bias, is a process where the scientists performing the research influence the results, in order to portray a certain outcome.
For example, drug research that is sponsored by the pharmaceutical industry, developer and producer of pharmaceuticals, is more likely to end up favoring the drug under consideration than studies sponsored by government grants or charitable organizations;evidence suggests (1) .
Cato Institute, public policy research organization, that "accepts no government funding" states in a published paper (2) fifteen specific practices of bias in the context of funded scientific research:
- Funding agency programs that have a biased focus.
- Agency Strategic Plans, RFPs, etc., with an agenda, not asking the right questions.
- Biased peer review of research proposals.
- Biased selection of research proposals by the agency program.
- Preference for modeling using biased assumptions.
- Biased peer review of journal articles and conference presentations.
- Biased meta-analysis of the scientific literature.
- Failure to report negative results.
- Manipulation of data to bias results.
- Refusing to share data with potential critics.
- Asserting conjectures as facts.
- False confidence in tentative findings.
- Exaggeration of the importance of findings by researchers and agencies.
- Amplification of exaggeration by the press.
- More funding with an agenda, building on the above, so the cycle repeats and builds
The study documents search terms for Google Scholar for each practice of bias that might be related to the government, example:
- Agency Strategic Plans, RFPs, etc., with an agenda, not asking the right questions.
Concept Analysis: Research proposals may be shaped by agency Strategic Plans and Requests for Proposals (RFP's), also called Funding Opportunity Announcements (FOA's). These documents often specify those scientific questions that the agency deems important, hence worthy of funding. Thus the resulting research proposals may be biased, speaking to what the agency claims is important rather than what the researcher thinks
Literature Snapshot: There appears to be very little research on this type of bias. A search on Google Scholar for the occurrence of the three terms "Federal" and "research" and "bias" in article titles for the period 2010-2014 gives zero hits! Full text search for these three common words returns over 62,000 items but only a few are even looking at bias, such as gender bias. The occurrence of all three words somewhere in an article is probably mostly just a coincidental combination.
Search for "research" and "bias" in titles gives 371 hits, but only a few of the articles are even related to federal funding. Most are looking at statistical analysis issues, such as sample selection.
And so on...
The study concludes that government funding has "received little" attention and has not been studied well; in other words the study didn't find any worthy empirical data regarding government funding and bias:
The role of government funding in inducing policy-driven bias seems to have received very little attention, even though it may be widespread. There are certain exceptions, most noticeably in the climate change debate and environmental policy in general. But here the attention is more a matter of public concern than one of quantitative scientific research. [...]
The notion of cascading systemic bias, induced by government funding, does not appear to have been much studied.
To dig deeper, some studies suggest that the question of bias of a research might be opinion-based:
It is important to remember that interpretation of data is inevitably subjective and can itself result in bias.
Conclusion: Apparently, there is little-to-no evidence regarding government research bias, cato.org reports. Governments got accused of crime research bias in the past, and professors wrote books accusing federal governments of having biased peer review systems since 1950, while other professors and notable scientists argue that the interpretation of data is inevitably subjective (3) :
Doctors are being encouraged to improve their critical appraisal skills to make better use of medical research. But when using these skills, it is important to remember that interpretation of data is inevitably subjective and can itself result in bias.
As skeptics, all we say is: "We don't really know".
(1) Als-nielson, B., W. Chen, C. Gluud, and L.L. Kjaergard. 2003. Association of funding and conclusions in randomized drug trials: A reflection of treatment effect or adverse events? Journal of the American Medical Association 290:921-928.
(2) Is the Government Buying Science or Support? A Framework Analysis of Federal Funding-induced Biases By David E. Wojick and Patrick J. Michaels April 30, 2015 CATO WORKING PAPER No. 29
(3) Kaptchuk TJ. Effect of interpretive bias on research evidence. BMJ : British Medical Journal. 2003;326(7404):1453-1455.
Per framework analysis by David E. Wojick and Patrick J. Michaels in April 2015, "the investigation of potential federally induced bias is not an active research area. There appears to be a major gap in bias research." The main example provided for government bias in the CATO working paper is regarding numerous allegations of bias linking federal policies on climate change and federal funding of climate research.
Conclusions and Observations: What we have found, as explained in this report, is that some types of bias are being studied extensively and quantitatively. Various aspects of peer review and publication bias, especially in biomedicine, appear to be the most heavily researched types of bias.
The role of funding in inducing bias is frequently alluded to as a potential financial conflict of interest. But it is not the focus of most research, which tends to look more at the practice of bias than at its cause. A new research thrust is likely needed.
Those who pay the piper call the tune (or at least get some influence on the playlist top 10)
The question isn't whether government rigidly controls what research scientists do but whether government has influence on it. And, unless you accept the improbable idea that all government is entirely beneficent and interested only in the truth, that influence creates bias. Showing that some scientists disagree with the government agenda doesn’t count as strong evidence against this idea. What needs to be shown is simply that the mix of research that gets done is influenced by what governments want to see. It is the portfolio of research as a whole that is biased if this influence exists not every individual piece of research by every scientist.
To show the potential for government influence it is worth looking at the sources of research funding. To show that bias exists it is worth looking at some research topics where we can see how the influence has changed the portfolio.
But there are several problems is detecting bias and finding its source. Some may arise from scientists not because of bias in government intent. Researchers in search of grant money may choose topics likely to be attractive to the grant-awarding bodies because they perceive this will enhance the possibility of getting future grants. Or the grant-awarding decision makers might influence researchers by promising future grants if the results are favourable (nobody has a problem believing this is true for industrial funders of research and the suspicion that it may be true is the reasoner strong disclosure rules around conflict of interest in many journals).
Another issue is that neither government nor science is a uniform, monolithic group. Individuals inside government may pursue a research agenda to influence other parts of government (this biases science even though the whole of the government machine doesn't have a single view). Scientists who have been appointed by government to grant-awarding bodies may pursue personal agendas. Is their appointment a sign of government bias or a personal issue?
Research Funding
Governments often fund a large proportion of research conducted within an economy, though this didn’t used to be true and is a legacy of two world wars. A contrarian view of this history is told by Terence Kealey.
In the USA, federal funding is the largest source of money for University R&D with more than $25billion spent per year according to this 2009 report. The UK spends a lot less (as a much smaller economy and less research intensive) but the government still funds the majority of university R&D and the overwhelming majority of R&D in public research institutions according to this ONS document.
So the preponderance of funding for “independent” research comes from government. We don’t expect industrial R&D to be independent of “bias” as its entire purpose is to pursue the interests of the sponsor.
So does government funding influence the content of that research? It seems obvious that it will unless government if regarded as spectacularly virtuous. But what about specific examples?
Examples of bias caused by government
The most obvious example where the agenda of science was not just biased but totally perverted is Lysenkoism in the Soviet Union under Stalin. As Wikipedia has it:
Ultimately, Lysenkoism failed to deliver on its promises in agricultural yields and even had unfortunate consequences such as the arresting, firing, or execution of 3,000 biologists due to attempts from Lysenko to suppress opposition to his theory, the president of the Agriculture Academy was sent to prison and died there, and overall, scientific research in genetics was effectively destroyed until the death of Stalin in 1953. Due to Lysenkoism, crop yields in the USSR declined as well.
Nobody expects democratic governments to resort to those sorts of methods to drive their agenda. But the funding still can have a strong influence on the research that gets done.
A topical recent example is the congressional ban on funding for research into the statistics or science relating to the use of guns in the USA. The CDC (which would normally have an interest in researching ways of reducing harm from avoidable causes of death) has effectively been prevented from doing so becasue of NRA sponsored Congressional restrictions since 1996 (see this recent summary at Vox). The general issue has also been addressed on this site in this question: Did the National Rifle Association (NRA) block research into statistics related to gun control?
So this is a negative way the agenda influences the portfolio of things that get researched.
But there are more subtle ways conventional and government opinion can influence the mix of research that gets done. An example is the way the harms of passive smoking have been researched. The conventional wisdom is that it must be bad (note that few will now question that smoking kills as the evidence is unusually overwhelming). But whether passive smoking is just annoying or seriously damaging to its victims is a lot less clear. Maybe. But it is almost unimaginable that public health authorities would tolerate research that says the effect is unclear. When Enstrom and Kabat published research pointing out the weak evidence base for the harms of passive smoking in the BMJ they were viciously attacked by the establishment.
The US government tried to prosecute them for their views (Wikipedia):
In 2006, prosecutors in a federal racketeering case filed documents which stated that Enstrom had received $94,500 from the tobacco industry between 1992 and 1997. The following year, the judge in this case, Gladys Kessler, ruled that major tobacco companies were guilty of racketeering and misleading the public regarding the dangers of second-hand smoke, citing the paper co-authored by Enstrom in the BMJ as evidence of this. ... In 2010, the University of California, Los Angeles School of Public Health announced that it would not be rehiring Enstrom because it felt his research was "not aligned with the academic mission" of their department. In 2012, Enstrom filed a lawsuit in federal court against UCLA in response to them terminating his position there. In the suit, Enstrom said that UCLA administrators "discriminated against Dr. Enstrom based on his ideological and political affiliations and sought to purge an academic dissenter from their ranks." In 2015, the case was settled, with UCLA allowing Enstrom to use the title "retired researcher" and continue to access university resources.
But Enstrom had always disclosed his funding (and, more importantly, probably could never have expected government funding for research leading to this conclusion). What is remarkable is the degree of irrationality in the original criticism of his paper. As a summary of the BMJ responses said:
Many letters were highly charged and hostile. “It is astounding how much of the criticism springs from Ad Hominem argument rather than from scientific criticism of the study itself,” wrote a “private citizen” from Philadelphia PA. “As a publisher of the leading Austrian medical online news service I feel quite embarrassed following the debate on this article. Many postings look more like a witch hunt than a scientific debate,” wrote another. It got bitter, and at times personal. A great read for anyone who enjoys a scrap. Disappointing for readers looking for a dispassionate appraisal of Enstrom and Kabat's study and its implications.
My point is that the establishment in general and government in particular reacted in a hostile way to contrarian views to the detriment of science. Enstrom’s views may be wrong (and some subsequent evidence suggests it was) but the response was to try and devalue and suppress the results rather than to do better research.
Another example where the government in the UK has tried to influence the research being done relates to policy on illegal drugs. David Nutt was the chairman of a supposedly independent group, the Advisory Committee on the Misuse of Drugs, that was supposed to influence government policy and has a statutory role in drug classification in the UK. Nutt was sacked from his role for publishing work suggesting that some popular drugs were substantially less harmful that the government believed. Many other members of the committee resigned in protest at this political interference.
Nutt has subsequently argued that government policy (especially the desire to send strong signals to conservative voters that they are taking a hard line on illegal drugs) has distorted many areas of useful research. For example he argued in Nature that:
An important and unfortunate outcome of the controls placed on these and other psychoactive drugs is that they make research into their mechanisms of action and potential therapeutic uses — for example, in depression and post-traumatic stress disorder — difficult and in many cases almost impossible.
Which is another way of saying that government has strongly influenced what sort of research gets done. In other words, when legislation makes research hard and government is unlikely to want to pay for research that disagrees with their stated policy, then the agenda for researchers can be tightly restricted.
Summary
When government doesn't pay for all the work done by scientists there is clearly scope for many to do things government doesn't like. Even this, though, is hard when the laws put barriers to certain types of research. The more subtle idea of policy driven evidence also causes bias.
Government typically has control of the majority of research funding. There are certainly instances when it uses this control and even legal and legislative means to influence the research agenda sometimes in ways that are not driven by a disinterested scientific agenda. This demonstrates bias. How big or significant this is to science as a whole is an open question, but it clearly exists.
What constitutes bias? How might be go about finding whether there's bias?
If someone wants to do health research in the US then it's important for scientists to qualify either for industry funding or NIH grants. For this question it doesn't make sense to look at bias caused by industry funding.
If those people who get grants don't find the truth but something else you could call the bias due to grant seeking government indused bias.
We have to therefore ask ourselves, whether such systematic biases exist in science or whether the scientific community manages to kill those biases. Before editing the original questions contained the assumption that the scientific process leads to truth. I therefore consider that to be a fair test of whether bias exists.
The Journal of Personality and Social Psychology published a paper called "Feeling the Future: Experimental Evidence for Anomalous Retroactive Influences on Cognition and Affect" by Daryl Bem in 2011 because the paper followed provided evidence for precognition that is as strong as the evidence in an average paper in the journal.
You don't need to be right to get a result published in academia. Precognition goes against the scientific consensus and therefore the paper got critized. A lot of papers who have similar evidence for the claims that they are making however don't go against the scientific consensus and therefore don't get critisied.
Breakthrough findings in basic cancer research seem to have a replication rate of 11%.
In a famous peer reviewed article titled "Why Most Published Research Findings Are False" John Ioannnidis comes to the conclusion:
Moreover, for many current scientific fields, claimed research findings may often be simply accurate measures of the prevailing bias."
Both cancer research and psychology are fields where it's easy to relicate experiments and make predictions about the outcome of experiments. Climate Research is a field where a lot of researchers think that successfully predicting the past with their model is as good as predicting the future. It therefore plausible that Climate Research isn't much better than cancer research or psychology at correcting against bias.
In addition to small scale grants, sometimes researchers also want to get the government to pay for something that's more expansive. As one example biologists wanted 3 billion dollar to sequence the human genome. If you would ask an average well educated person about the percentage of the human DNA that got sequenced I would guess that most people without a background in biology would say 100%. In reality the number is something like 92%. Those biologists had an incentive to show the general public that the 3 billion dollar were well spent money. They profit if the public believes that they actually archived more than they achieved.
Similarly if you read the fine print of the IPCC report, the report concludes that "net effect of human activities since 1750 has been one of warming" is a claim with "At least 9 out of 10 chance of being correct". In the public discourse it's hard to tell people that a 9/10 of being true should motivate people to reduce emissions. Many people can't handle it intellectually that a claim isn't either true or false but can be 90% true. As a result a lot of the public communication is targeted at getting people to believe that the claim has a 10/10 chance of being true. That what people do who want to achieve the politically goal of lowering emissions.
It's problematic because if we had models that would 100% accurately predict what the climate does geoengieering would probably outperform cutting emissions. Explaining to the public that a model is good enough to advocate cutting emissions but not good enough for geoengieering is hard. It's politics.
