Politically selective calls for open access
SPARC Open Access Newsletter, issue #151
November 2, 2010
by Peter Suber
What should we think about politically selective calls for OA?  For now, put aside those that are yoked to general calls for OA and framed as politically realistic first steps.  What about those that are not yoked to general calls for OA, and whose narrowness suggests political opportunism more than political realism?  Here are five quick examples from the US to show what I have in mind.

(1) When Bill Clinton ran for President in 1992, Republican critics called for public disclosure of Hillary Clinton's undergraduate thesis.  When Barack Obama ran for president 16 years later, Republican critics called for Michelle Obama's undergraduate thesis.  When Obama nominated Elena Kagan to the Supreme Court, Republican critics called for Kagan's undergraduate thesis.  As far as I know, none of the people calling for public disclosure of these undergraduate theses called for similar disclosure of the theses written by the spouses of Republican presidential candidates or by the Supreme Court nominees of Republican presidents.  Nor did they call for OA to all undergraduate theses. 

(2) In 1999, Senator Richard Shelby sponsored an amendment to an appropriations bill, subsequently passed, requiring federally-funded research projects to make their data publicly available through the Freedom of Information Act.  (Green OA mandates were unknown in 1999.)  Shelby opposed certain environmental regulations and wanted the studies underlying those regulations exposed to skeptical scrutiny, including some studies based on private medical records of patients harmed by airborne pollution.  But his FOIA-access mandate was not limited to those particular studies or even to environmental research.  On the contrary, the Shelby amendment was a general open data mandate and applied to publicly-funded studies of all kinds.  Eleven years later, however, in February 2010, Senator James Inhofe cited the Shelby amendment as one ground for criminal prosecution of climate scientists unwilling to share their data.  One of the targeted scientists was Michael Mann, creator of the famous "hockey stick" graph showing a sharp rise in global temperatures in the 20th century.  Like Shelby, Inhofe opposed certain environmental policies and hoped that public scrutiny of the underlying studies would undermine the policies.  (In a public statement as famous as the hockey stick graph, Inhofe called the idea of human-made global warming the "greatest hoax ever perpetrated on the American people.")  Unlike Shelby, however, Inhofe was unwilling to generalize the call for access, even to climate research, and has a track record of opposing generalized OA policies such as the NIH policy. 

(3) In 2005, Representative Joe Barton led a House committee investigation of three climate scientists, including Michael Mann.  Barton, who denied the role of human activity in global warming, gave the three scientists 18 days to produce documentation on hundreds of studies going back decades.  When scientists and members of Congress criticized Barton's investigation as harassment, an anonymous blogger defended the investigation and claimed to speak for Barton's committee.  The blogger cited arguments on the benefits of OA to research in general, although neither Barton's committee nor the blogger were calling for OA to research in general.

(4) In 2006, liberal and conservative members of the House and Senate agreed to co-sponsor a bill creating an OA database of US government spending.  Initially the bill was general and applied equally to liberal and conservative projects.  Each side was willing to gamble that public exposure would help its cause more than the opposing cause.  However, the House bill was soon amended to favor conservatives by exempting the disclosure of government contracts with corporations.  In the Senate, the unamended general bill was blocked by two Senators, one conservative and one liberal, who wanted to limit public scrutiny of the pork projects they steered toward their states.  Eventually the balanced, general bill passed both chambers and became law.w

(5) Earlier in 2010, Kenneth Cuccinelli II, the Attorney General of Virginia, picked up where Joe Barton (#3) and James Inhofe (#2) left off, and investigated climate scientist Michael Mann for fraud.  Like Barton and Inhofe, Cuccinelli believes that human-made global warming is a hoax, and targeted one of the most prominent scientists opposing his views.  Like Barton, he demanded a mountain of documentation, going back over a decade, and gave Mann's university just over a month to produce it.  Like them both, he has not called for OA to climate research generally or OA to publicly-funded research generally.  In August 2010, a federal court halted his investigation.  But early last month, he picked it up in a modified form that he hopes will pass muster with federal courts.  About two weeks ago, the University of Virginia was back in court to stop him.  Stay tuned.

See the appendix for links and more detail on these five cases.

* These cases beg for a close analysis of the kind of opportunism that puts political advantage over principle.  If these politicians thought that climate science had become ideologically skewed, if they wanted to improve its reliability, and if they believed public scrutiny would help expose and correct scientific errors, then why did they target just a few political opponents when they could have taken a more general, principled approach?  Why not call for OA generally, to all research in the field, to help expose and correct errors generally, across the field?  Why target just the leading scientists opposed to their own personal views, as if their own personal views were the measure of scientific reliability, or as if their own personal views were the remedy, rather than the recipe, for ideological bias?

Here, however, I'm more interested in the strategy questions these tactics raise for OA activists, and by extension for activists of all kinds.  When should we welcome small steps toward a good goal and when should we oppose them?  Does it matter whether the small steps are politically selective if they take steps in the right direction?  Does it matter whether there are political obstacles to wider, fairer, and more general policies? 

First I acknowledge that all five cases show Republicans putting political advantage over principle.  One case (#4) shows a Democrat joining a Republican in this practice.  I've looked for other examples of Democrats doing the same, on OA issues, without much to show for it.  The next closest case I've found is that some Democrats joined some Republicans in opposing OA for the high-quality, publicly-funded research from the Congressional Research Service.  Politicians of both parties have called for OA to CRS Reports (notably, John McCain and Joe Lieberman) and politicians of both parties have opposed OA to the reports (notably, Bob Ney and John Larson).  http://www.earlham.edu/~peters/fos/2007/12/bill-to-mandate-oa-for-crs-reports.html

If you can help balance the picture, drop me a line.  I don't want this investigation to be guilty of one kind of bias while criticizing another. 

But for the issues I care about here, it doesn't matter whether the guilty parties are Republicans or Democrats, conservatives or liberals.  The issue is whether we should support politically selective calls for OA, on the ground that they would take at least a small step in the right direction, or whether we should reject them because of their motivation, their incompleteness, their political consequences, or some other failing.

Some of the five cases don't even take a step in the right direction.  For example, I can't support mandatory OA to undergraduate theses (#1) or patient medical records (#2), although I support voluntary OA to both when the students or patients consent.  But I strongly support mandatory OA for publicly-funded research, including climate research and data (##2, 3, 5), and mandatory OA for data on government spending (#4). 

Even after we sort out the cases, however, and distinguish those taking steps toward goals we can support, we still face a hard question.  When should we support incremental progress, especially when it's politically lopsided, and when should we tell our opportunistic, one-time-only allies to return to the drawing board?

* These cases bother me because in most respects I'm an incrementalist.  I even wrote an article in 2004 in defense of incrementalism:

One reason to praise forward steps is that they really do constitute progress, whether or not they reach full OA....There's no contradiction here and no implication that the progress we are praising is all that could be desired.  We praise students for improving without implying that they've improved enough.  We praise good moves in politics, sports, and science without implying that further progress is unnecessary or impossible.  We can speak clearly.  If we want to say that widening access is good and open access is better [or that targeted OA is good and general OA is better], we can say that. 

I was very disappointed in 2004 when the NIH proposed an encouragement policy with an embargo up to 12 months, especially when Congress had asked for a mandate with an embargo no longer than six months.  But I was an incrementalist and argued that we should support both the imperfect policy and continuing strategies to strengthen it. 

When the East Anglia "climategate" case broke, a common response across the scientific world was to call for OA to climate data.  I was heartened when politicians joined scientists in calling for open data, including some who had already decided on nonscientific grounds that the scientific consensus must be wrong.  I didn't support the cynical wing of the coalition, but I did support the general call that the cynical had cynically joined.  I supported it even though the call didn't go beyond climate science to all science or even all publicly-funded science.

When emails stolen from climate researchers at the University of East Anglia's Climatic Research Unit suggested that some researchers might have fudged some data, scientists of many kinds called for open data for all climate studies.  Because the scientists calling for open data were joined by some serious politicians concerned to keep climate science credible with the public, and by some opportunistic politicians who don't follow and don't accept climate science, the odds rose that public policy on climate research might shift toward OA....

When we agree that a goal is good, then we should do what we can when we can.  If we have an opportunity to open up climate data today, but no similar opportunity for national security data, then we should open up climate data today without waiting until we can do everything at once.  If some climate activists want to open up climate data and then stop, we should support them with the intention of continuing to work when they themselves have finished. 

Here's the hard question:  If we have an opportunity to open up climate data from a few targed researchers, before we have an opportunity to open it up more generally, should we take it?  Here we need to ask some follow-up questions.  (1) Will we open up the kinds of data we really want open?  For example, will we open up ocean temperature data, rather than personal emails about ocean temperature data?  (2) Will we allow the kinds of exceptions or delays we really think are legitimate?  For example, will we exempt non-anonymized medical data and wait until the data-gathering scientists have had a chance to publish their results at least once?  (3) Is it true that we can't yet open up data more generally?  For example, are we targeting just a few individuals because we really can't do better, or would it be just as politically feasible to call for OA to all research of a certain kind? 

If we can answer 'yes' to all three questions, then let's take the step.  But if the step would open up information not useful for research, the case for OA is weak, not strong.  If the step would harm researchers or research subjects, then the cost could easily outweigh the benefit.  If a larger and more general step is just as feasible as the smaller step, then limiting ourselves to the small step is a missed opportunity for progress and a needless narrowing of focus requiring --but usually not receiving-- a special justification.

Here's an analogy.  Suppose we support efforts to reduce carbon emissions.  If a politician pretended to agree and proposed "as a first step" to ban all carbon emissions from certain people, all targeted political opponents, that would be indefensible.  The problem is not that it fails to qualify as a small step toward a good goal.  It does qualify.  The problem is that it invidiously discriminates on political grounds, creates harms to offset the good, and chooses a narrow focus when a general rule would be more desirable, more justified, and apparently as feasible.

To pretend that general arguments for reducing carbon emissions justify this selective step is to overlook the most objectionable aspect of this selective step.  Any justification for the step must address the objection about its suspiciously political, narrow scope.  This was the problem with the anonymous blogger's defense of Barton's committee investigation in case #3.  The general arguments for OA were impeccable, but they didn't answer the objection that the investigation was unjustifiably narrow and political.

The problem is not with intentions but consequences.  The hypothetical carbon ban --like the actual Barton-Inhofe-Cuccinelli investigations-- may be intentional political warfare.  But partisan intentions needn't disqualify a bipartisan plan, which is one of the heartening aspects of case #4.  When assessing a proposal we can ignore intentions and look only at the proposal itself.  In the hypothetical case, however, it's hard to deny that the small first step could have been a tolerable carbon reduction for a fairly selected group rather than an intolerable burden for an unfairly selected group. 

What if we actually support the political goals of a politician who happens to use such a one-sided tactic?  The question is not whether we support the players or causes that might gain from the tactic, but whether we can support the tactic itself as a means to the end.  If it creates political gains for causes we like, we can't overlook its costs to innocent bystanders and fair procedure.  (Adlai Stevenson comes to mind:  "The hardest thing about any political campaign is how to win without proving that you are unworthy of winning.")

Sometimes small steps are positively preferable to large ones, even when both are politically feasible.  For example, we could reduce risk in off-shore oil drilling if we decide to study the consequences of small, early steps before committing to larger, later ones.  But this principle doesn't cover Barton, Inhofe, and Cuccinelli, since they've offered no public plans to generalize the call for OA after studying the consequences of the politically selective OA they demand.

The question in these cases is when small, politically selective steps toward OA are acceptable even when they are *not* preferable to larger, general OA policies.

The best reason to accept small steps, when they're not preferable to large steps, is that sometimes they are politically feasible when larger steps are not.  For political reasons, the 2005 version of the NIH policy stopped at mere request and encouragement.  But it was better than the previous non-policy, which was a reason to support the advance.  However, stopping short of a mandate was also a reason to keep working for a mandate.  The 2008 version of the policy strengthened the request to a requirement but failed to shorten the embargo.  It was better than the previous voluntary policy, which was a reason to support the advance.  However, failing to shorten the long embargo was also a reason to keep working for a shorter embargo.  

We've had to work toward a mandatory policy with a short embargo in stages, not because small steps were better than large ones but because accomplishing everything in one step was politically impossible.  Likewise, if we want libre OA mandates rather than gratis OA mandates, then we must still work for them in stages.  The political obstacles are precisely why gratis OA policies are hard-won, and worth celebrating, even if we regard them as steps along the way to libre OA policies.

In short, to repeat, when the goal is good, do what you can when you can.  Sometimes you can only do so much.  But turn the ratchet of social good whenever you have your hand on the crank. 

When large steps face insuperable political obstacles, it's much easier to support small steps.  But when larger steps are politically feasible, and certain smaller steps would have one-sided political consequences, then we have to think about how to weigh the stepwise progress against the risk of playing an invidious political game.

Barton, Inhofe, and Cuccinelli seem to think that public scrutiny of Michael Mann's data (and personal emails) would invalidate his results.  But they know very well that the vast majority of scientists in the field support his results, and have already independent data to come to similar conclusions.  In fact, the existence of that scientific consensus has put these politicians on the defensive and provoked their political strategy.  But what is their strategy?  If making Mann's work OA would only tend to confirm it in the eyes of other scientists, then why would *they* want to do it?  It's hard to avoid the conclusion that their real purpose is harassment and intimidation.  They want to raise the cost to scientists who confirm inconvenient truths about climate change.  If raising the cost and annoyance level is success, then they're succeeding. 

That's why I think we can go beyond talk about intentions to talk about consequences.  If they intended to expose bad science, we wouldn't have to agree with them about which science is bad; we should focus on the consequences and welcome their support in the campaign to share research data for every sort of review and reuse.  If they intended to harass and failed, then we needn't care.  But if they intended to harass and succeeded, then we should certainly care.  They may not be stopping other climate scientists from following the weight of the evidence, but they are increasing the cost of doing so --and ironically all in the name of OA.

It's just as hard to avoid the conclusion that another element in their strategy is FUD.  We start from the same premise:  these politicians are very aware of the scientific consensus on climate change.  They must realize that exposing Mann's data would do little or nothing to shake that consensus.  But in their demands to expose it, the lay public will see the consensus contradicted, even if it's contradicted by politicians rather than scientists.  Politicians who can't disconfirm a scientific theory can still shake its credibility, at least with the constituents that matter most to them.  If they reduce its credibility and increase FUD, they'll reduce the odds that policy-makers and the public will accept large and expensive solutions that would inconvenience the same constituents.  For those politicians, that's at least half a victory. 

When the East Anglia climate researchers were acquitted of misconduct by the Commons Science and Technology Committee of the British Parliament', "the MPs criticised Professor [Phil] Jones and climate scientists in general for being too possessive and secretive about the raw scientific data and computer codes they use to establish the link between global warming and human activities. They also criticised the [University of East Anglia] for fostering a culture of non-disclosure of scientific information to climate sceptics."

That secrecy, plus political attacks on climate science and scientists, plus superficial journalism dutifully giving equal weight to both sides of a cooked-up controversy, have had a cumulative effect.  The credibility of climate science has taken a hit, way out of proportion to its apparently normal rate of methodological problems.  Worse, it's impossible to say that some of that hit wasn't deliberately solicited or abetted by political operatives and PR firms who specialize in sowing doubt.  Actual misconduct in climate science may be rare, but suspicion and distrust are not.  On the other side, OA can answer suspicion and distrust, at least when the OA is general and not politically selective.  On this front, general OA policies are part of the solution and politically selective calls for OA are part of the problem.

Among the many groups calling for general OA policies as part of the solution was the Independent Assessment Panel set up to evaluate the climate data controversy.  From its April 2010 report:

It was pointed out that since UK government adopted a policy that resulted in charging for access to data sets collected by government agencies, other countries have followed suit impeding the flow of processed and raw data to and between researchers. This is unfortunate and seems inconsistent with policies of open access to data promoted elsewhere in government.

The University of East Anglia was certainly not alone in "impeding the flow of processed and raw data to and between researchers."  Not even Al Gore and the publicly-funded UN Intergovernment Panel on Climate Change made their work OA, not even after they made public statements about the importance of spreading the message they were publishing, and not even after they had Nobel prize money to compensate them for any lost royalties. 

But the generality of the problem calls for general solutions.  Even when the selective targets of calls for OA are guilty of needless secrecy, our solutions should be general.  When the selective targets are guilty of nothing more than disagreeing with policy-makers, or ratcheting up the inconvenience of inconvenient truths, then we should redouble our support for general solutions and not lend our weight to intimidation or FUD, even when they're dressed up as small steps toward OA.  To advance the cause --and incidentally prove our good faith, if that's necessary after refusing to join certain politically selective steps toward OA-- let's work for OA to climate research in general, or publicly-funded research in general, or peer-reviewed research in general. 

Let's do that with our eyes wide open to the fact that good knowledge doesn't lead directly to good policy.  We must work on both parts of that two-part problem.  Let's use general OA policies to advance good knowledge, and use the resulting good knowledge to advance good policy.  Let's learn what we can and use what we know, even if it's foreseeable that we'll get bogged down later quarreling about how to translate our good knowledge into good policy.  Let's not cooperate with attempts to set policy without using what we know, attempts to limit the circulation of knowledge, attempts to intimidate those trying to know, or attempts to engineer FUD and confuse what we know.

In particular cases, our decisions will turn on fine judgments about how large a step is politically feasible or how damaging a certain narrower step would be.  We won't all draw the line of unacceptable incrementalism at the same place.  True friends of OA disagree about whether to support double-dipping hybrid OA journals, whether to support any hybrid OA journals at all, or whether to support OA journals as such, or gold OA, until green OA is further along.  They disagree about whether gratis OA is an acceptable incremental step toward libre OA, whether CC-BY-NC is an acceptable incremental step toward CC-BY, and whether CC-BY is an acceptable incremental step toward CC0.

The same is true on the other side as TA publishers experiment with OA or decide which incremental steps they can support.  In June 2006, the Association of Learned and Professional Society Publishers (ALPSP) and the International Association of Scientific, Technical & Medical Publishers (STM) issued a joint statement in support of open data.  I blogged this comment at the time:

The ALPSP and STM both lobby *against* policies that would provide OA to research *literature*, like FRPAA and the draft RCUK policy. I acknowledge that there are many differences between OA to data and OA to peer-reviewed articles interpreting or analyzing data.  But ALPSP and STM should acknowledge that there are many similarities, and that most of their arguments for OA data (enhancing research productivity, avoiding costly repetition of research, supporting the creative integration and reworking of research) also apply to OA literature.

Similarly, some TA journals provide OA at their own expense to especially important articles or editorials, in order to spread the word.  But that creates line-drawing problems when they must decide when a work is sufficiently important to qualify for special treatment.  Some TA journals provide retroactive OA to articles that turn out to be especially important, for example when their authors later win Nobel prizes.  (There are some examples in this month's Roundup section, below.)  But that creates the same kind of line-drawing problems, even when there are generalizable ways of doing the same thing.

The negative spin on these practices is that some TA publishers recognize the principle of OA but apply it selectively or even inconsistently.  The positive spin is that these TA publishers recognize the principle and are taking incremental steps to live up to it.

In any hard case of incrementalism, one question is about the pace of progress.  But another is about changing minds.  Part of any activist campaign is to change minds, and part of changing minds is to move from one position to another incompatible with it.  If we're too harsh on the inconsistencies of others, whether in allies or opponents, then we'll stigmatize and deter the kind of mind-changing that is our goal.  But if we're more forgiving, in order to open the door to the kind of progress we want, then at least we should understand that we too are being politically selective. 

* Here are some cases to test your own intuitions about where to draw the line. 

Last spring the Russian Federal Archives Agency began providing OA to papers documenting the 1940 Soviet execution of Polish officers at Katyn.  The documentation is coming out in dribs and drabs, but this is apparently due to archiving problems within the agency, not political interference.  I applaud the project, but since I don't know the cause of the problems, perhaps I should be more conditional:  I'm ready to applaud the project if the plan is for general disclosure, without political selectivity, and the problems are merely logistical.

In May 2010, the Washington State Supreme Court ruled that public libraries may use internet filters to block pornography.  At least filters don't violate the free-speech rights of patrons, and librarians are free to decide how much of the unfiltered internet to offer patrons.  Some librarians wanted to take the big step of offering the whole unfiltered internet.  But Jan Walsh of the Washington State Library wanted the state's public libraries to take the smaller step of offering a porn-free internet.  She argued:  "If you want to see the demise of support for libraries, just keep going the way of wide-open access to all of these ugly sites."

Librarians continue to disagree about which position best fulfills the mission of libraries.  I suspect that nearly all librarians would agree to take the larger step, and offer the unfiltered internet, when it is politically feasible to do so, or when there are politically feasible ways to implement it within the library, for example, by lifting filters on request for adult patrons in certain parts of the library.  Making unfiltered access lawful doesn't solve the political problem in communities where library policies can cost them patrons and funding.

In an April 2010 letter to BMJ from Ane Krag Jakobsen and four co-authors pointed out that fee-based OA journals "preferentially increase[] accessibility to studies funded by industry [which makes funds available for publication fees]. This could favour dissemination of pro-industry results...."

My take:  this is a small step toward a good goal --not the goal of industrial domination of science but the goal of financially healthy peer-reviewed OA journals as an alternative to peer-reveiwed TA journals.  Insofar as industrial sponsorship favors some conclusions over others, it's a one-sided step.  But it should be complemented by other steps rather than retracted.  It's not invidious that industrial sponsors are willing to pay publication fees at fee-based OA journals.  Indeed, every kind of research funder should do the same, when it can afford to do so.  We should increase support from funding agencies and universities, which we're doing, rather than reduce support from industry.  Meantime, as always, we should push for green OA, which requires no fees, for all peer-reviewed research regardless of how it is funded. 

Until more non-industrial sponsors enter the lists, or until green OA levels rise further, industry-sponsored research may (or may not) be more accessible than non-industry-sponsored research in fee-based OA journals, and that may (or may not) skew some investigations.  But this kind of bias is peculiar and deserves at least two comments.  First, non-industrial sponsors are joining industrial sponsors in funding OA journals and may already outnumber them.  In that sense, the correctives are politically feasible and spontaneously growing.  There may be no threshold at which they become knowably sufficient, but they're moving steadily in that direction.  Second, the bias will only affect researchers who limit their reading to fee-based OA journals, a very odd research method that I've never heard anyone propose or defend.  Moreover, it would be difficult to put into practice, deliberately or inadvertently.  Not only are OA journals a minority of peer-reviewed journals, but fee-based OA journals are a minority of OA journals.  Most readers don't know and can't easily tell whether an OA journal charges author-side publication fees.  If the bias were more likely or more severe, then we'd want to offset it without waiting for more corrective steps to accumulate.  But because ordinary research practices steer clear of the problem, we can easily tolerate this first step while subsequent steps steadily improve the solution.  Indeed, we should praise industrial research sponsors for their commitment to OA.

Also see thoughtful comments on the Jakobsen letter by Kent Anderson and S. Pelech.

However, similar kinds of bias have actually arisen and done harm.  Ellen Roche died in June 2001 when she inhaled hexamethonium for an asthma experiment.  The physician supervising the experiment researched the safety of inhaling hexamethonium, but apparently limited his research to one contemporary textbook and PubMed.  Some journal articles from the 1950's documented the risk of death from inhaling the drug, but PubMed's coverage didn't start until the 1960's.

Here the problem is not any bias in the literature, but the limited scope of one OA resource, a side-effect of incrementalism.  That incrementalism doesn't always mix well with the unprecedented convenience of OA, which makes OA resources the first stop for many researchers, and for some busy or careless ones, the last stop as well.  This cases raises some deep questions about incrementalism, and I addressed some of them in this newsletter for August 2001:

...Fourth, I'd urge researchers to draw the conclusion that relevant information may exist without being online, let alone free and online.  When easy searches fail, we do not have a bona fide negative result from which we can draw scientific conclusions.  Instead we must spend the time and calories to undertake a more arduous search.  Fifth, I'd wonder out loud whether there is a good scientist anywhere who didn't already know this.  Finally, I'd point out that part of the underlying problem here is that [OA] is compellingly attractive.  It is only hazardous because it tempts busy people to rely on it to the exclusion of other research methods.  What makes it tempting is not false advertising but spectacular convenience....If [OA] is spectacularly convenient, but not yet adequate [in its scope or coverage], then it may be hazardous to put before people who don't understand its inadequacy or whose will has been weakened by various pressures.  But the same is true of potato chips and money.

If the problem is that [OA] is available for use before it is adequate, then the real objection is to stepwise progress.  Which is better, to make [OA] repositories available as they are ready, and let them grow and interconnect in real time, or to hold them back until the network of such repositories encompasses all published literature?  Serious scientists would be the first to object to the latter plan.  Running a close second would be people with debilitating medical conditions anxious for research breakthroughs....

Let's also admit that PubMed doesn't pretend to be more than it is.  Its scope does not go back to the 1950's or extend to the rulings of federal administrative agencies [which showed that the US Food and Drug Administration withdrew its approval of hexamethonium in 1972].  What it does, it does very well....If it appears to a researcher through a veil of illusion, the illusion of sufficiency, then it is the researcher's illusion.  The same is true of a print library.

If incomplete resources lead to inadequate searches, with fatal consequences, then we might want to question incrementalism itself.  But we can't realistically demand that researchers search only complete resources.  No resources are complete and the full interoperability of our many incomplete resources is still a long way off.  This is another way back to the conclusion that the problem lies more in the inadequacy of the researcher's search than the inadequacy of any incrementally growing resource or library.

One more example:  In 2009, the Reynolds Tobacco Company went to court to demand a copy of an unfinished book manuscript by Stanford historian, Robert Proctor.  Proctor had often served as an expert witness against tobacco companies and was scheduled to do so again in an upcoming trial.  Reynolds wanted to use information from Proctor's book to prepare its cross-examination of him in that trial. 

Although Reynolds was not demanding OA for Proctor's manuscript, an important OA principle is still at stake here.  We want OA for texts that authors are ready to publish or distribute, but not before.  Violating that principle leads to Congressional subpoenas for personal emails, as if to advance research.  In other circumstances, violating the principle could force disclosure of patentable discoveries before authors have a chance to apply for patents.  That could mobilize the very large and well-funded patent industry to oppose OA policies they could otherwise support, which is roughly what happened with the 2003 Public Access to Science Act (a.k.a. the Sabo bill).

Reynolds was not demanding OA to all unpublished book manuscripts, or even private disclosure of all unpublished book manuscripts.  But I won't criticize the company for failing to generalize, for those generalizations would have been worse than the selective demand, which was already bad enough.  Moreover, while Reynold may not have generalized its demand, US courts treat decisions as precedents for similarly situated parties in the future, and in that sense stand ready to generalize where Reynolds wasn't.  The generalization would have been something like this:  unpublished book manuscripts by expert witnesses should be disclosed privately to the parties against whom the witnesses plan to testify, when those parties need access to the manuscripts in order to prepare their cross-examination of those witnesses.

* Appendix.  Here are some more details, links, and comments on the five examples from the first section.  I repeat a little from my summaries there to make these case studies stand on their own.

(1') When Bill Clinton ran for President in 1992, some Republican opponents called for public disclosure of Hillary Clinton's allegedly radical undergraduate thesis at Wellesley College on Saul Allinsky.

When Barack Obama ran for President in 2008, some Republican opponents called for public disclosure of Michelle Obama's allegedly radical undergraduate thesis at Princeton on whether a Princeton education made Black students more motivated, or less motivated, to help the Black community.

The Obama campaign made Michelle's thesis OA after the election. 

In response to similar calls in 2010, the Obama administration also released Elena Kagan's undergraduate Princeton thesis on socialism, after Obama nominated Kagan to the Supreme Court and conservative critics demanded to see her thesis.

As far as I know, Democrat critics of Republican presidential candidates have never called for public disclosure of the undergraduate theses of the candidates or their spouses.  As far as I know, none of the Republicans calling for disclosure of the Clinton, Obama, and Kagan theses called for the public disclosure of the undergraduate theses of the spouses of Republican presidential candidates, or the nominees of Republican presidents to the Supreme Court.  Nor did they make a general call for OA to all undergraduate theses. 

I support OA for doctoral dissertations, even mandatory OA for doctoral dissertations.

But while I support voluntary OA to undergraduate essays, I couldn't support an OA mandate.  Many of the strongest arguments for OA to doctoral dissertations do not carry over to undergraduate theses.  Undergraduate theses are valuable as pedagogical exercises even when they fall short of the criteria used to judge dissertations.  Teachers may ask for original scholarship in undergraduate theses, but are more likely to get it in doctoral dissertations, which makes dissertations better subjects for inclusion in the institution's publish-or-perish policy.  Mandatory OA is a way to treat a dissertation seriously as a work of original scholarship in which the university takes pride.  Doctoral students are trying to become professionals in a field, and can be asked to take the risk of publishing something they may regret later. 

But even if I were willing to call for mandatory OA to undergraduate theses, and even if I knew that Hillary Clinton, Michelle Obama, and Elena Kagan all wrote admirable theses (which would not be hard to believe), I'd still have to conclude that people targeting these three theses were thinking more about embarrassing political opponents than sharing valuable scholarship.

(2') In 1999, Senator Richard Shelby (R-AL) introduced an amendment to a spending bill requiring that the Office for Management and Budget (OMB) revise one of its regulations to mandate that "all data produced under an award will be made available to the public through the procedures established under the Freedom of Information Act (FOIA)." 

Remember that this was 1999, when there were no green OA mandates anywhere.  Requiring a certain category of publicly-funded data to be available through FOIA was a precursor to today's OA mandates.  If Shelby had known about green OA mandates, he might have rewritten his amendment to create one.

For details, see the 2002 report on the Shelby amendment from the US National Academies Press (NAP), "Access to Research Data in the 21st Century: An Ongoing Dialogue Among Interested Parties Report of a Workshop."

Digression for OA historians:  Chapter 5 of the NAP report acknowledges the clumsiness of the FOIA for data sharing and considers some alternatives to streamline the process, including free online access to the datasets and the software required to read the data.

The Shelby amendment passed.  This was a breakthrough for open data arising from publicly-funded research.  So what's the problem?

There's no problem except Shelby's motivation, and even his motivation wasn't a large problem in light of the generality of his amendment.  According to the NAP report, the Environmental Protection Agency started regulating airborne particulate pollution in 1997, based on studies that it caused health problems and its levels correlated with death rates.  Quoting the report (pp. vii-viii):

Critics of the proposed standards claimed that implementing the new standards would be unreasonably costly, with estimates reaching billions of dollars. They argued further that the standards were not scientifically justified, and they called for access to all of the underlying data so that the results could be verified by other scientists. The Harvard researchers [who undertook one of the targeted studies], who had been funded by the National Institutes of Health (NIH) and not the EPA, declined to provide the underlying data. They took this position in large part because of their concern that releasing all data would violate the privacy agreements they had made with the patients who participated in the study as part of the informed consent process. However, Harvard did indicate a willingness to provide the data to other qualified researchers for confidential analysis and requested that the Health Effects Institute (HEI) conduct an independent review of the data.  The researchers’ refusal to make all their data available led to calls from Congress, industry, and others, requesting access to the data, and was one of the causes of...the Shelby Amendment...."

Shelby opposed the pollution regulations and wanted to scrutinize the data from the studies inspiring the regulations in order to undermine them.  However, to his credit, wrote a general OA (or FOIA) mandate that went well beyond the studies underlying the regulations he disliked.  His political allies may have shared his skepticism about the science, or his priorities for business over the environment, but they too used general arguments for data sharing.  From the NAP report (p. viii):

Defenders of the Shelby Amendment argued that it provided the public with both accountability (taxpayers fund the research —therefore they should be able to see its basis) and transparency (the public should be able to review research data produced with federal funds that is used to support regulatory decisions that affect the public).

For that reason, Shelby's personal motivation doesn't matter much.  He may have opposed environmental research and some environmental regulation, but his amendment would accelerate and improve research of all kinds.

But while I support his generalied call for open data, I also respect the data-sharing reservations of the scientistis behind the studies that especially agitated him.  Those studies were based on medical data showing the effects of airborne particulate pollution on individual patients.  If the researchers obtained those data on the condition that they would keep them private, they were bound to keep them private.  Moreover, I support the principle that medical privacy takes priority over OA, or that only anonymized medical data can be made OA.

If that were the end of the story, I wouldn't use it as a case study, except perhaps as a case study showing that a one-sided motivation is compatible with a general OA policy.  But the Shelby amendment returned to the news in February 2010 when Senate Republicans released a minority report on the "climategate" controversy from November 2009.  Although the notorious emails were leaked from the Climatic Research Unit of the UK's University of East Anglia, one of the researchers involved was an American, Michael Mann of the University of Virginia, which created a foothold for a US Senate committee to investigate. 

The Senate minority report threatened criminal charges against Mann for falsifying data, and cited the Shelby amendment as one of the laws he might have violated.  The report also alleges (p. 29) that Mann might have violated the Freedom of Information Act (FOIA), a directive from the White House Office for Science and Technology Policy, President Obama's Transparency and Open Government Policy, the Federal False Statements Act, the False Claims Act, and federal laws against the obstruction of justice  and iInterference with Congressional proceedings.

The report was written by Senator James Inhofe (R-OK), in his role as the ranking minority member of the US Senate Committee on Environment and Public Works.  Inhofe famously called global warming the "greatest hoax ever perpetrated on the American people" (July 2003).

Unlike Shelby, Inhofe was not willing to generalize.  Inhofe has not called for OA to all climate data, or all publicly-funded climate data.  He even tried his best to derail the OA policy at the NIH before its adoption.  In October 2007, he filed two amendments to the Senate appropriations bill containing the OA mandate for the NIH.  One amendment would have deleted the NIH provision and the other would have weakened it significantly.  His strategy was apparently to use the deletion proposal to set up the dilution proposal as a reasonable compromise. 

Charles Bailey pointed out at the time that one of Inhofe's top financial contributors was Reed Elsevier.

(3') In 2005, Joe Barton (R-TX), then-chairman of the House of Representatives Energy and Commerce Committee, began investigating three government-funded scientists --Michael Mann, Raymond Bradley, and Malcolm Hughes-- who concluded that human activity is largely responsible for global warming.  Barton denied their conclusions, worked for the oil industry before running for Congress, and in the decade before his committee investigation was one of the top five Congressional recipients of oil-industry campaign funds.  He gave the three scientists 18 days to produce documentation on hundreds of studies going back decades.  Scientists and fellow members of Congress criticized his investigation as harassment. 

In 2006, an anonymous blogger who claimed to be writing on behalf of Barton's committee began using OA arguments to defend Barton's investigation.  The posts have since been taken offline, but I blogged excerpts at the time:

Climate change is a fascinating science worthy of much study. Some recents [sic] studies have been used by overzealous regulators and politicians to push heavy-duty burdens and taxes on many industries. Before we tax potentially trillions of dollars out of the economy, we here at the House Energy and Commerce Committee thought we might have a look at it too. Turns out, that made us personna [sic] non grata.

Now, the National Research Council, in a report that upholds the science that hypothesizes on some some [sic] warming trends, also upholds our efforts to look at the data too.  As you can see on page 23 of the report's overview section, the NRC took note of the issue of access to scientific data, and emphasized the importance of sharing information.

"Our view is that research benefits from full and open access to published datasets and that a clear explanation of analytical methods is mandatory," the report said. "Paleoclimate research would benefit if individual researchers, professional societies, journal editors, and funding agencies continued to improve their efforts to ensure that these existing open access practices are followed. "

Obviously, our nation's most prestigious scientific [sic] sees the need to make data available.

I accept the National Research Council's arguments for open data, especially in high-stakes and contested fields like climate science.  The difference between the NRC's call for OA and Joe Barton's was that the NRC was willing to generalize.  It called for a policy to cover a whole field, or to cover all publicly-funded science.  Barton zeroed in on three scientists whose influential results contradicted his own beliefs. 

It turns out that the three scientists had been reluctant to share their data, not just with Barton but with scientific colleagues.  Some of the data, Mann argued, was "protected by confidentiality agreements with governments."  (This is similar to the confidentiality problem that caused the data-withholding that inspired the Shelby Amendment in case #2.)

But if we want to open up research data not covered by confidentiality agreements, should we target three researchers or call for a more general opening up? 

One more complexity:  Joe Barton may not have called for OA to climate research generally, or publicly-funded research generally, but he did support the NIH policy and did use his committee chairmanship to help it pass. 

I applaud him for that, and wish he'd appeal to the same policy considerations behind the NIH policy to generalize his interest in public scrutiny for climate research.  In this respect Richard Shelby is a good model.  A desire to undermine certain studies by exposing them to scrutiny is entirely compatible with a general OA mandate. 

(4') In 2006, a liberal Democrat (Barack Obama, D-IL) and conservative Republican (Tom Coburn, R-OK) co-sponsored the Federal Funding Accountability and Transparency Act of 2006, which would create an OA database of US government spending. 

At first the proposal was general, disclosing the pet projects of both liberals and conservatives.  And at first Senators from both parties supported it, though their motives might have been one-sided and complementary.  Liberals liked it because they thought public scrutiny would protect popular programs and the politicians who funded them.  Convervatives liked it because they thought public scrutiny would expose needless programs and protect politicians who wanted to axe them.  That looked like democracy in action:  people who think government programs are justified, and people who think they are unjustified, agreeing to back their bets with public scrutiny and accept the public's verdict. 

The bill's only opponents were politicians who wanted to shield their pork projects from public scrutiny.  But over time that small category grew large, especially on the Republican side.  House Republicans won amendments which required the disclosure of grants, which benefitted nonprofit organizations favored by liberals, and blocked the disclosure of contracts, which benefitted corporations favored by conservatives.  In the Senate, two Senators notorious for steering pork projects to their own states, Republican Ted Stevens and Democrat Robert Byrd, put secret "holds" on the bill, stopping its progress, but eventually relented.

Eventually the balanced version of bill passed the Senate by unanimous consent, and the House adopted the Senate version.  The OA database, USASpending.gov, went live in December 2007. 

The bill started with admirable generality, veered toward politically selective OA, and then regained admirable generality with strong bipartisan support.  It's a success story.

If the only politically feasible version of the bill were the selective version that suppressed information about government contracts with corporations, should supporters of OA government data support the bill, in the spirit of starting a long journey with a single step, or would that be a case in which the politics swamps the principle?

(5')  In March 2010, Kenneth Cuccinelli II, the Attorney General of Virginia, picked up where Joe Barton (#2) and James Inhofe (#2) left off, and investigated climate scientist Michael Mann for fraud.  (Mann worked at the University of Virginia 1999-2005, though he now works at Pennsylvania State University.)  Cuccinelli is a conservative Republican who "really believes that global warming is a hoax" according to Jeremy Mayer, professor of public policy at George Mason University.  Cuccinelli says that his investigation is to defend Virginia taxpayers by discovering whether Mann used doctored research to seek research funds from the state.  According to the Chronicle of Higher Education, "Cuccinelli gave the university just over a month to produce what could amount to thousands of documents." 

[T]he scope of Mr. Cuccinelli's demand for documents from the University of Virginia was nothing short of breathtaking. The "civil investigative demand" delivered to the university seeks "data, materials, and communications that Dr. Mann created, presented, or made" in connection with five grants involving a total of about $484,875. But it also seeks "all documents that constitute or are in any way related to correspondence, messages, or e-mails" between Mr. Mann and 39 other scholars, including many prominent climate scientists, as well as his communications with "all research assistants, secretaries, or administrative staff" with whom he worked at Virginia.  For good measure, it demands "any and all e-mails or pieces of correspondence from or to Dr. Mann since he left the University of Virginia," and "any and all computer algorithms, programs, source code, or the like created or edited" by Mr. Mann. The demand covers documents going back to January 1, 1998—three years before Mr. Mann gained international notice as co-creator of the "hockey stick": a graph showing that global temperatures, after remaining stable for several hundred years, rose sharply in the 20th century.

As far as I know, Cuccinelli has never called for OA to climate research generally or to publicly-funded research generally.  Even if Cuccinelli did support general calls for OA, like the NIH policy and FRPAA, there are reasons why those policies do not cover private emails, just as there are reasons why they omit undergraduate theses. 

The University of Virginia went to court to dismiss Cuccinelli's demand for documents, and in August 2010 it prevailed.

However, in early October Cuccinelli revised and reissued his demand for documents, and in late October the University of Virginia went back to court to stop him.  (It was the revival of Cuccinelli's investigation that made me decide to address this topic this month.)


Read this issue online

SOAN is published and sponsored by the Scholarly Publishing and Academic Resources Coalition (SPARC).

Additional support is provided by Data Conversion Laboratory (DCL), experts in converting research documents to XML.


This is the SPARC Open Access Newsletter (ISSN 1546-7821), written by Peter Suber and published by SPARC.  The views I express in this newsletter are my own and do not necessarily reflect those of SPARC or other sponsors.

To unsubscribe, send any message (from the subscribed address) to <SPARC-OANews-off@arl.org>.

Please feel free to forward any issue of the newsletter to interested colleagues.  If you are reading a forwarded copy, see the instructions for subscribing at either of the next two sites below.

SPARC home page for the Open Access Newsletter and Open Access Forum

Peter Suber's page of related information, including the newsletter editorial position

Newsletter, archived back issues

Forum, archived postings

Conferences Related to the Open Access Movement

Timeline of the Open Access Movement

Open Access Overview

Open Access News blog

Peter Suber

SOAN is licensed under a Creative Commons Attribution 3.0 United States License.

Return to the Newsletter archive