In order to implement the US Office of Science and Technology Policy (OSTP) public access mandate, most of the federal agencies involved will have to go through an elaborate rulemaking process. Publishers and others who are potentially impacted by this open access (OA) mandate might start thinking about how to participate in these rulemakings. To that end here is a brief strategic overview, apologies in advance for all the acronyms, but government seems very fond of its TLAs and FLAs (Three and Four Letter Acronyms)

To begin with each agency has to do its own rulemaking, but there may be ways to combine them administratively into one or more multi-agency rulemakings. So we may be looking at a lot of separate agency actions, each of which has to be addressed in its own way, or just a few. We really do not know at this point how burdensome and complex the process will be. This is the first “monster” that I described in my “Three Monsters and a Gorilla” article on the OSTP mandate. Organizations with specific concerns might want to set up small coalitions to coordinate and simplify their participation.

The core of every rulemaking is the publication of a proposed rule and the submission of comments thereon. Bloggers think commenting is new but federal agencies have been taking comments on proposed actions for many decades. There may also be one or public hearings on the agency proposals. Note that the so-called comment period is often quite short, typically just 30 to 90 days. Thus it is good to develop comments ahead of the proposal publication, to the extent possible.

Interestingly the National Research Council has already held hearings on the OSTP mandate, as well as accepting written comment submissions. Many of these submissions read like formal rulemaking comments. This activity was sponsored by the affected federal agencies but it is unclear whether these comments will be part of the rulemakings to come. This is important because if someone wants to sue an agency for bad rulemaking, their suit generally must be based on the written record, which may not include the NRC submissions.

Because of the specialized nature of the OA rules, there may also be some important additional administrative procedures. These involve the Office of Information and Regulatory Affairs (OIRA) in the Office of Management and Budget (OMB). OIRA (pronounced oh-eye-ruh) enforces two administrative mechanisms which are designed to promote dialog and to moderate agency actions. As explained below these are the Paperwork Reduction Act and the long standing Regulatory Review executive order. See also OIRA’s FAQ page:

Under the Paperwork Reduction Act (PRA) every agency information collection must be approved or “cleared” by OIRA. The agency OA systems mandated by OSTP are clearly information collections as defined by PRA. Any reporting requirement is an information collection under PRA. In this case authors or publishers are reporting the publication of articles that flow from agency funding.

OIRA approval is via what is called the clearance process, which is designed to give maximum stakeholder input, which OIRA takes seriously. OIRA often acts as the arbiter between the agency involved and an affected industry, such as the publishers in this case.

The clearance process involves two rounds of public notice and comment, which are separate from the rulemaking described above. First the agency or agencies involved must publish their proposed information collection plan and take public comments on it. They then prepare an Information Collection Request (ICR) and submit it to OIRA. This ICR submission is also published for comments, which go to OIRA. OIRA may, and often does, return the ICR to the submitting agency for revision based on the comments it receives.

A key part of the ICR is a burden estimate. This is the agency’s estimate of how much work its collection will require, usually expressed in annual hours. OA systems can be quite burdensome, for example if authors are required to submit manuscripts to the agency. Every federal agency has a “burden budget” or a total allowable burden and new ICRs are governed by these limits so there is great pressure to minimize burden. (Disclosure: I helped set up OIRA many years ago.)

Then there is the long standing Regulatory Review executive order. This requires what is called “regulatory review” by OIRA. If an agency mandates an OA system this is a form of regulation, making it potentially subject to prior regulatory review by OIRA. One of the criteria for regulatory review is adverse economic impact on a specific business sector and studies indicate that OA mandates seem to fit this case. On the other hand the usual dollar threshold for regulatory review is probably larger than the potential economic impact of the OA rules. So whether or not a proposed OA system will trigger regulatory review is itself an open, and potentially controversial, question.

Under Regulatory Review the agency must prepare a Regulatory Impact Assessment (RIA) for the proposed action, in this case its OA system mandate. The RIA must include consideration of alternative regulatory system designs plus a cost benefit analysis. Doing cost benefit analysis of federal OA proposals is an intriguing prospect. Also interesting is the fact that OIRA is organized by agency, so these many OA rulemakings will cut across much of its staff structure. Then too OIRA makes federal information policy, which might include OA mandates. One wonders if OIRA knows this wave is coming at them.

In any case, this regulatory wave is certainly coming at the scholarly publishing industry. Best to prepare.


8 Thoughts on "US Federal Open Access Rulemakings To Come"

Note that if an agency were to rely entirely on CHORUS links to published content that should eliminate the need for a rulemaking, nor would it be an information collection under PRA. This is a big plus for CHORUS.

I managed the websites for a major US Government science agency. From this perspective, I’m afraid I have zero confidence in any attempt to create an agency-wide repository of papers.

The problem? Placing their published papers in an agency repository, or even just reporting the titles and abstracts, is an additional burden to busy scientists. My agency could not even successfully maintain a list of published articles. Our compilations, dependent on authors for input, were hopelessly incomplete. Agency regulations required published scientists to submit their information, but there was just no way anyone was going to impose serious and morale-killing discipline.

The scientists, for their part, would say, “What’s the problem? My article is in a reputable journal and well indexed. Anyone who needs to can find it. Why spend more resources?” That’s a hard argument to counter, especially since many agencies are going to Gold OA.

PubMedCentral seems to be doing pretty well as far as compliance goes. If FundRef catches on it will make enforcement much easier. Mind you NIH has specific statutory authority, something the other funding agencies lack.

PubMed Central has big stick to back it up. Want more NIH grant money? Make sure your pubs are there! Other science agencies, NOAA, USGS, etc. do a lot of work in-house.

Ken, are you suggesting that the other funding agencies are going to defy OSTP and refuse to implement public access systems? Anything is possible in Federal Fantasy-land but I have a hard time seeing this happen.

Nobody is going to openly defy a directive; that just doesn’t happen. Resistance by benign neglect is another matter. Agency budgets are tight and scientists have to prioritize their time.

It should probably be noted that a great deal of PMC’s success in compliance is due to journals voluntarily depositing articles on behalf of authors. This saves authors a tremendous amount of work and confusion, and is perhaps an area where journal publishers are not given adequate credit for helping out.

Good point. This is way more effective than having authors self report. Outside the medical field, however, there is no PubMed. It would be very critical for agencies to have a common reporting system, as I’m sure journals would not want to implement many different types of reports.

Comments are closed.