Open access policies must require rather than just request deposit



SUMMARY: It makes a big difference whether a university’s (or research funder’s) Open Access policy is mandatory (i.e. a requirement) or just invitational (a request). Merely requesting deposit (as did the first version of the NIH OA policy) fails to generate deposits, whereas requiring deposit succeeds (as did the revised NIH policy, once it was upgraded to make deposit mandatory). Research grant recipients and their institutions are conscientious about complying with their funders’ official grant fulfillment procedures, to maximize their chances of future funding. University faculty are likewise conscientious about complying with their institutions’ official performance review procedures (and so depositing articles in the institutional repository should be designated as the official procedure for submitting them for performance review).

Klaus Graf wrote in the American Scientist Open Access Forum:

KG: “request” or “require” is only a play on words.

So is “may” vs. “must.” And “recommended” vs. “obligatory.”

But words matter, when it comes to formulating official institutional policy.

And they matter all the more in an area that is still very new and unfamiliar to most researchers, hence still rife with confusion and misunderstanding, as is OA: “What’s in a Word?

KG: You cannot compare a funder mandate (NIH) with a university mandate.

You certainly can — and must, if you are to formulate effective OA policies. The two kinds of mandates are complementary: “How To Integrate University and Funder Open Access Mandates

KG: Request in a funder mandate means: “May be there will be disadvantages if I don’t selfarchive”

Funder mandates (like NIH) never explicitly specified the disadvantages of noncompliance, but during the two years that the NIH policy was just a request there was only 5% compliance whereas within a year of upgrading the policy to a requirement compliance exceeded 60%.

(The obvious disadvantage of noncompliance with a funder mandate is that grantees risk not receiving a future grant if they fail to meet their present grant’s official requirements — as opposed to doing the things that are merely “requested” or “recommended” or “optional.” The positive advantage of compliance is enhanced research impact.)

KG: Request in a university mandate means: “Nothing will happen if I do so”.

To get a more realistic idea of the contingencies, please have a look at those university mandates that are procedurally tied to the official mechanism for submitting articles for university performance review — for example, the U. Liege mandate, to which the Rector has already drawn your attention in a prior posting:

“Yesterday, Klaus Graf reacted rather strongly to the announcement of the Liège University repository mandate, stating [in the American Scientist Open Access Forum] that its ‘practice and legal framework is nonsense.’

“It seems to me that perhaps he may have missed a few essential aspects of this mandate, essentially the way it is handled in practice, the legal wherewithal and the reasons for imposing it….

Excerpt from Liege Mandate: “…starting October 1st, 2009, only those references introduced in ORBi will be taken into consideration as the official list of publications accompanying any curriculum vitæ in all evaluation procedures ‘in house’ (designations, promotions, grant applications, etc.).”

(Hence the obvious disadvantage of noncompliance with a university deposit mandate is that if faculty fail to comply with their institution’s official submission procedure for research performance evaluation, their articles will fail to be evaluated. This is rather like a procedural requirement to submit a digital rather than a paper draft, or even a draft in a particular digital format. Note that the university also shares a stake in its faculty’s compliance with funder requirements, both the disadvantages of noncompliance and the advantages of compliance. The positive advantage of compliance in both cases is enhanced research impact. The disadvantage of noncompliance is loss of future grants, including both their research impact and their contribution to institutional overheads and indirect costs.)

KG: Harvard-style: “I can get all waivers I need”.

The jury is still out on what will prove to be the compliance rate with the Harvard-style mandates (with their option of opting out).

I have argued that the Harvard-style mandates should be upgraded to immediate-deposit mandates that allow authors to waive adopting the author’s addendum (on copyright retention and re-use rights), but not to waive making the deposit itself (for which access can be set as “Closed Access” if they wish to honor a publisher embargo period). Currently, deposit itself is not part of the Harvard mandate, just part of the accompanying Policy FAQ; but I still have hopes that the wording of the Harvard policy itself will be formally upgraded to make deposit mandatory in all cases, rather than contingent on whether or not the author opts to waive adopting the author addendum:

Harvard Mandate Adds ID/OA to its FAQ
Which Green OA Mandate Is Optimal?

KG: I cannot see any proof that the very few documented high deposit rates after a mandate have the mandate as causa instead of the readiness of a faculty/university to deposit.

Well it would be a remarkable coincidence indeed if the difference between the (many, many) institutional repositories with low deposit rates (<15%) and the very few that have high deposit rates (>60%) were the fact that the latter happen to have faculty with a “readiness to deposit” — rather than the more obvious difference, which is that they require deposit!

(If the real causal difference is a local “readiness to deposit” rather than the official requirement to deposit, perhaps we should be looking at what the faculty are eating at those universities, so we can add it to the diet of the faculty at all those other universities whose faculty do not yet seem to have this estimable “readiness to deposit”…)



Response to an early posting by Klaus Graf (about peer review, PLoS, which publishes biological and medical research, all of it peer-reviewed (fortunately for us all).

Preprints in monographs free online if they wish to. The trouble is that most do not wish to (yet). But all journal-article authors already do.

This issue has next to nothing to do with peer review.

Conference proceedings fall under the same category as journal articles (author give-aways, written solely for usage and impact).

Edited book chapters are an in-between area. The best strategy is to get all the journal articles safely and universally self-archived, and the rest will follow soon enough. Don’t get hung up on the exceptions and outliers.

KG: It is wrong to think that all relevant research is made from university affiliated scholars. It would be good to have valid numbers for scholars without deposit access to an institutional repository.

True again — but again, no point getting hung up on the exceptions and outliers: Get all institutionally generated research articles self-archived (85% still waiting!) and don’t worry about the exceptions and outliers for now.

But, yes, central repositories (like DEPOT — or CogPrints or Arxiv) are just fine for self-archiving institutionally unaffiliated research.

KG: Institutional repositories are NOT better than central disciplinary repositories.

Opinion duly registered.

KG: Repetition [does not] make… false things… true.

You can say that again…

Stevan Harnad
American Scientist Open Access Forum