The taxpayer argument for open access
SPARC Open Access Newsletter, issue #65
September 4, 2003
by Peter Suber
There are many arguments for open access.  Let's focus here on just one:  the argument that taxpayers have a right to open access to the results of the research funded by their taxes.

If the research papers based on taxpayer-funded research are locked away in conventional journals that require payments for access, then taxpayers will end up paying twice for the same research.  The primary version of the argument is that it would be wrong to make taxpayers pay a second fee for access.  A secondary version of the argument is that tax money should be spent in the public interest, not to create intellectual property for the benefit of private publishers, who acquire it and profit from it without paying the authors or compensating the public treasury.  Both versions of the argument object that taxpayers are paying twice when they shouldn't have to.  The first version of the argument focuses on overpayment, while the second focuses on misuse of one or both payments.

A third form of the argument holds that the current U.S. rule to put the works of tax-paid government employees into the public domain (Section 105 of the Copyright Act) should be extended to the works of tax-paid government contractors and grantees.  It holds that the arguments for our current policy also apply to this extension of the policy, so that it would be inconsistent to support the current policy and oppose its extension.  I won't examine this third form of the argument here.  Other aspects of the argument take all my space, and I dealt with some of these issues in SOAN for 7/4/03.

Here are five objections to the taxpayer argument and at least five replies.  My purpose is to clarify the argument, identify its strong and weak forms, and show where it comes to an end and must be supplemented by other arguments for open access.

(1) Objection:  Taxpayers can walk into a library that has paid for access, and read journal articles without paying to do so.  Or they can receive copies of the articles by interlibrary loan.  We already have free access by taxpayers to taxpayer-funded research, and to most other research as well.

Reply:  This is free public access to a paid copy.  The subscribing library had to pay retail to make this access possible, and to do so it had to forgo many other purchases.  Moreover, of course, the library payment is on top of the previous taxpayer payment.

So this kind of access is a red herring that does not answer the taxpayer argument or fill the need for open access.  In addition, most public libraries don't subscribe to a good range of scientific and scholarly research journals, and most academic libraries (according to W. Wayt Gibbs) are cutting back the access privileges they accord to "walk-in" patrons not affiliated with the institution.

W. Wayt Gibbs, "Public Not Welcome: Libraries cut off access to scientific literature", _Scientific American_, August 11, 2003.
(Full-text accessible only to subscribers.)

(2) Objection:  Open access to federally funded research will only affect a subset of the scientific and scholarly journal literature.  Federal funding supports the natural sciences much more than the social sciences and humanities, and some natural sciences much more than others.  Why is open access to one field's results more important than open access to another field's results?

Reply:  The taxpayer argument is that *at least* the government-funded research should be open access, not that *only* this research should be open access.  It's about fairness to taxpayers, not fairness to the disciplines.  Moreover, there's no harm in solving a large problem one step at a time, especially if different steps have different justifications.  Finally, there are many arguments for open access, and some of them (as we all know) apply to all the disciplines even if the taxpayer argument does not.

The taxpayer argument does not intrinsically distinguish among the disciplines, and does not imply that some are more important than others.  Instead, it applies to any research that is funded by taxes.  It's contingent what research falls into this category and the taxpayer argument, by itself, is indifferent to the policies that determine which disciplines and research projects get funded.  If tax dollars only funded research in one narrow specialization, such as the history of the umbrella, then the taxpayer argument would only require open access to that narrow band of research.  Conversely, if all research were funded by taxes, then the taxpayer argument would require open access to all of it.

(3) Objection:  When the government gives a research grant, it is funding the research itself, and perhaps the writing of articles based on that research.  But the journal is charging for value it adds to that research and writing:  peer review, copy editing, manuscript preparation, marketing, and publishing.  So while subscription fees are additional payments, they don't pay for the same thing as the government research grant.

Reply:  This is true.  If the taxpayer argument is that federal research grants and journal subscription fees pay for the very same product, then it's mistaken.  Some who speak loosely might make this mistake.  But most who use the argument take it in a slightly different direction.  They argue that the primary value of a journal article lies in the research and writing.  When taxpayers have already paid for this primary value, then their access to the resulting journal article should not be held hostage to secondary expenses, at least when some of these are unnecessary, some are overpriced, and the rest can be subsidized so that they needn't become access barriers for readers.

Note three particular aspects of this refinement of the argument.

First, even if all its assertions are true, it depends on open-access arguments (such as the adequacy of upfront funding) that go beyond the taxpayer argument.  In that sense the taxpayer argument is limited in scope and must be combined with other open-access arguments to take us all the way to the conclusion.

Second, it concedes that journals add value to the author's research and writing.  I only pause to point this out because publishers often overlook the fact that open-access proponents agree with them about this.  Open-access proponents might even concede that journals add value in every way that publishers say they do.  However, open-access proponents tend to argue that some of these journal services are more essential than others, even if all are valuable, that they cost less than most journals charge for them, that it's better to cover the costs of the essential services from the author's end of the transaction than the reader's end, and therefore that they do not justify access barriers.  Open-access proponents and commercial publishers may never agree on these four propositions.  But debating them takes us well beyond the taxpayer argument.

Third, when the taxpayer argument is applied to archives rather than journals, then it doesn't face the same limitation and doesn't need the support of other open-access arguments.  Open-access archives don't perform peer review, copy editing, manuscript preparation, marketing, or publishing.  Their expenses are tiny compared to those of journals.  For archives, there is virtually nothing else to pay for once the taxpayers have paid for the research and writing of an article.  The software for hosting an OA archive is open-source, the labor for installing it is negligible and donated, the server space is inexpensive and easily justified by the archive's benefits to the hosting institution, and the labor of uploading new content is distributed among the participating authors.  Since archives are a true form of open access, this means that the taxpayer argument does suffice for at least one form of open access, even if it only takes us halfway to the other.

Sometimes the taxpayer argument is restated so that it is about ownership rather than access.  The Public Library of Science used to argue that parents should pay the midwife but then keep the baby, or that the midwife should accept payment but then return the baby.  The analogous taxpayer argument is that taxpayers *own* the research they fund, and that the midwife-publisher should be paid for services rendered, not for access by the owner to what it owns.  Moreover, the midwife can be fully compensated without transferring title to the baby.  (I always liked the midwife analogy and am sorry that PLoS seems to have stopped using it.)

When refined to avoid an obvious mistake, then, the taxpayer argument concedes that subscription fees and government research grants do not pay for the same goods.  Hence, it doesn't object that taxpayers pay twice for the same goods.  It objects that one payment ought to suffice --and, under a better publishing system, would suffice.  The second payment isn't a duplicate of the first, but either a needless access barrier (because the expenses it covers can recouped in other ways) or an improper one (because it gives midwives improper control of the baby, and private-sector enterprises improper control of a public investment).

Note how refining the argument has led us to shift from its primary form (about avoiding duplicate payments) to its secondary form (about spending tax money in the public interest).

This version of the argument combines nicely with the argument I'm not examining here, that we should extend our current policy on in-house government research to government-funded research.  The taxpayer argument shows that taxpayers have made the primary investment in this body of research, and perhaps own it.  The copyright statute shows that as a nation we have already decided that when taxpayers have this relationship to a body of research, then it's not private property for private profit.

(4) Objection:  At best, the taxpayer argument supports open access for taxpayers, not open access for the whole world.  If access can be open for one nation and closed to other nations, then that's as far as this argument compels us to go.  In fact, if we could fine-tune access even further, and limit it to those citizens who actually paid taxes and deny it to fellow citizens who didn't, then the taxpayer argument would not stop us.

Reply:  This is a fascinating objection whose strength depends on a complicated combination of other government policies and empirical facts about profit and loss.  In short, if enforcing this access barrier costs less than it brings in, then the taxpayer argument has no objection to it; but otherwise, the taxpayer argument requires that we remove the barrier and provide open access to the whole world.

Leave aside the problem that "open access" to one country and not others isn't really "open access".  The question is what the taxpayer argument implies, not what the term "open access" implies.

First note how this logic has been applied in Canada.  The Canadian government supports the National Research Council (NRC), which supports the NRC Research Press, which publishes 15 peer-reviewed journals.  On January 1, 2001, Canada adopted the policy that Canadians should have free online access to these 15 journals, while citizens of other countries would have to pay.  The free access for Canadians is funded by the Depository Services Program, a public-private partnership funded in part by the Canadian Treasury Board.

I wish I knew whether the 15 NRC journals only publish research funded by Canadian taxpayers, whether the DSP has arranged the free online access by buying it from the NRC Press at retail prices, whether all or only some of the funding for Canadian access comes from Canadian taxpayers, and whether the cost of limiting open access to Canadians is offset by subscriptions from non-Canadians.  Unfortunately, I don't know any of these things.

But what's interesting is that the taxpayer argument seems to entail no more than the nation-limited kind of open access that Canada has arranged.  Moreover, there seem to be good reasons for the limitation.  If the argument is that the taxpayers who funded certain research shouldn't pay twice, then it aims to protect those taxpayers, not foreigners and others who didn't pay to fund the research.  If the argument is that tax money should be spent in the public interest, then it probably applies only to the national interest.  The world interest may trump the national interest in ethics, but tax policy is a peculiar domain in which taxpayers can have good reasons to favor themselves over other and larger populations.

Against these arguments is the simple fact that it *costs less* to provide unrestricted access to all internet users than to discriminate between authorized and unauthorized users and block access to the unauthorized.  If taxpayers deserve open access, then they deserve it without paying extra for the apparatus to block others.

Behind this we see the familiar reasons why open-access journals have lower expenses than conventional journals.  They dispense with subscription management, password registration, and authentication filters.

Why should Canadians pay for open access by other countries?  Here's an analogy.  An insurance company might buy a radio advertisement that it knows will be broadcast to a cluster of states, including a state where it does no business.  If the technology existed to block the transmission to states where it does no business, the company would have no interest in paying extra to use it.  The company's interest is in saving money and getting its message to a certain audience.  From its point of view, over-distribution is harmless, and only under-distribution is harmful.  If over-distribution costs less than precise distribution, then it's a bargain.  This line of reasoning should carry even more weight for nations funding open access to research papers, which are useful across all boundaries.

So this branch of the taxpayer argument --for worldwide rather than merely national access-- reduces to the venerable one that, other things being equal, the government should take the lowest acceptable bid.  It should not pay more than necessary for the goods that policy has decided the public should have.  This direct, practical argument means that we don't even have to reach the more nuanced, ethical argument that we should always prefer to buy two goods for the price of one, when we can, even if one of the goods will be enjoyed by people other than ourselves.  But the second argument is as valid as the first.  We should not spitefully deny others a costless benefit or a benefit that costs no more than what we would have spent on ourselves.

But what if the cost isn't the same?  What if the access-limiting apparatus paid for itself in subscription fees from non-citizens?  That would mean that ordinary bean counting would entail nation-limited access, without a whiff of nationalism or spite.

When access-limiting apparatus pays for itself in subscription fees, then it does nullify the cost argument for worldwide access.  In fact, it would even reverse the cost argument insofar as this apparatus would then be a means for reducing the cost of national open access.

Canadians who want free access to the 15 NRC journals must go to a certain web site and register their IP address and ISP.  Presumably software at the NRC journals then checks user IP addresses against this national registry.  This is expensive compared to putting the same articles on the web without all this armor.  Because Canadians could get open access without this extra investment, we should ask what the extra investment brings them.  The answer is "nothing at all" --unless subscriptions bought by non-Canadians reduce the price that Canadians pay for their own open access.

A country that paid for its own open access through fees charged to outsiders would be putting economic efficiency ahead of the public good from public knowledge.  This might run afoul of other policy-based arguments for open access, but we must acknowledge that it would not run afoul of the taxpayer argument.

Canada's National Research Council (NRC)

NRC Research Press journals

Canadian government's Depository Services Program (DSP)

Web form for Canadians to fill out to get free online access through the DSP to the NRC Research Press journals

FAQ on free access for Canadians

Press release announcing the free online access for Canadians, June 18, 2000 (the free access began January 1, 2001)

(5) Objection:  Ordinary taxpayers don't need to read peer-reviewed scientific literature and wouldn't understand it if they did.  Researchers would benefit from access to this literature, but they form a small subset of taxpayers.

Reply:  This may be true (more below), but it's beside the point.  The taxpayer argument doesn't say that taxpayers should have open access because they need it, or because it would be useful to them; it says they should have open access because they've paid for it.  If you buy a house, you should be allowed to enter, even if someone can argue that you didn't really need to buy it.

Of course other open-access arguments assert that open access is useful.  Indeed the usefulness or public good argument is the main argument.  It is so common and so central that it makes the version of the taxpayer argument emphasizing buyer's privilege or buyer's rights a kind of anomaly.   It's important to remember this when focusing on the taxpayer argument, which in many ways takes us in a direction away from most other open-access arguments.  We must also remember that even the argument from usefulness or the public good needn't assert that open access is equally useful to every kind of person or that everything useful ought to be free.

When Martin Sabo and some PLoS members use the taxpayer argument, they emphasize the value of open access for individual patients suffering from serious diseases, or the value of open access for relatives and friends browsing on their behalf.  This is a weak form of the argument that invites the present objection.  Unfortunately it often crowds out a much stronger version of the argument that could have been used in its place.

This form of the argument is weak because, in fact, most individuals don't need to read peer-reviewed medical literature and wouldn't be able to understand it.  It's even weaker if it puts more weight on the emotional drama of one person's story rather than the evidence added by that story to the total case for the conclusion.  It's a judgment call whether a given rendering of the argument crosses this fine line.

Rick Weiss opened his August 5 story in the _Washington Post_ with an anecdote about an ill child in a poor family.  In the following days, participants on the SSP discussion list ripped Weiss for making such a bald appeal to emotion.  Quoting one post:  "I find it disgusting and reprehensible that certain proponents of open access are preying on the fears of families and holding out the false hope that cures for devastating chronic illness may lie in the pages of hard-to-find medical journals."  The problem is not that open-access proponents are opening themselves to this kind of criticism, or that the SSP list is predominantly pro-publisher and anti-OA.  The problem is that for some open-access rhetoric, the criticism is justified.

The argument from individual sufferers is only a fallacious appeal to hope and fear if it puts more than an anecdote's worth of weight on the emotional anecdotes it tells.  But it needn't be fallacious at all.  There are individuals without medical training who can understand some peer-reviewed medical literature and who benefit from reading it first-hand, and it's undeniably true that open access will help realize this benefit.

I am one such person.  I've used online medical research (some open access, some through the licenses bought by my college) for myself and for close relatives.  I've used it for serious illnesses and for a slew of minor questions ranging from diet to fitness.  I often learn more from this first-hand reading than I learned from my family doctor, and often explore issues that I wouldn't bother to raise with my doctor.  I'm grateful for the access that made this research and exploration possible.  The point is not that I'm rare, but that this benefit is small compared to the benefit of open access *for researchers*.  It's small even if there are a lot of people like me who benefit from doing their own online research.  It's small even if we place a great value, as I do, on the benefits of first-hand reading and research.

The reason is simply that open access by medical researchers will help everyone by accelerating the progress of medicine, while open access by individual sufferers will only help a much smaller number of people and in much less significant ways.  Both benefits are real.  But when making the case for open access, let's lead with the strongest version of the argument, not the most sentimental.  We don't have to be silent about other, lesser benefits.  But we invite misunderstanding and criticism when we lead with the lesser benefits and leave the impression that there are no larger ones.

Rick Weiss's August 5 article in the _Washington Post_

The SSP list (from the AAP)
(The list archive is accessible only to subscribers.)

To read the SSP list archive, first subscribe to the list.  Then go to <>.  Under anonymous access, enter "SSP" as the list, and your email address.  Then click "Enter NTList".  Once logged in, click "Archives".  (Believe it or not.)

* Conclusion.  The taxpayer argument can be misleading.  It seems to say that government research grants and journal subscription fees duplicate one another.  But they don't and a careful form of the argument will acknowledge this.  It seems to imply that open access is required only for taxpayers, not the world, but this depends on some empirical contingencies that will differ from place to place and time to time.  It can locate the benefit of open access in individual taxpayers who suddenly gain access to peer-reviewed medical literature, or it can locate it in the gain to everyone when open access accelerates the progress of medicine and the other sciences.  When stripped of mistaken assertions and disentangled from other arguments for open access, the argument can answer the objections raised against it.  It is stronger when cast as an argument about spending tax money in the public interest than when cast as an argument against duplicate payments or overspending.  It is stronger still when combined with other arguments for open access.  On its own, it is stronger for open-access archives than open-access journals.  When combined with other arguments, it may be equally strong for both, depending on how you judge the other, non-taxpayer arguments for open access.

* Postscript.  My analysis supports the following conclusions about the taxpayer argument made by others.

From the August 6 public statement by ARL and SPARC supporting open access to federally funded research:  "[T]he purpose of federally funded research is to advance knowledge and accelerate new discoveries benefiting society at large. It is critical that this new knowledge be readily available to physicians, researchers, and individual members of the public, including those who are unaffiliated with or remote from libraries that subscribe to increasingly expensive journals and databases that are the current outlets for much federally funded research. Broad and timely open access to the results of this research is essential to maximize the return on taxpayer investment."

From a July article in _Serials_ by Jan Velterop:  "The argument that publicly funded research should be publicly available is a valid one, but it cannot be the most important one in the discussion as to whether research should be freely accessible. The overriding argument is that freely accessible research optimises the scientific process as well as its 'translation' into societal benefits."
(Accessible only to subscribers.)

From a June 26 press release by the Susan G. Komen Breast Cancer Foundation:  "[P]roviding [open] access to research ­ with both positive and negative findings ­ can be instructive for individual health decisions and strategies, as well as for the construct of further research."

From a June 26 press release by the Public Library of Science:  "PLoS argues that the current closed system of scientific publication places the narrow interests of publishers before the public interest and greatly diminishes the value of the more than $50 billion dollars invested by US taxpayers each year in scientific and medical research."


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 <>.

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