The Minister of Higher Education and Training published an update of the Policy for Measurement of Research Outputs of Public Higher Education Institutions (2003), now called the Research Outputs Policy, 2015 on 11 March 2015.

The purpose of the new policy update is to renew and regulate the system whereby the government, through its Department of Higher Education and Training (DHET) subsidises the research output of public higher education institutions in South Africa.

It is our intention with these comments to both alert interested parties to important changes to the 2003 policy and its implications, and to give a personal perspective on the new policies from the viewpoint of a scholarly publisher in South Africa. After all, the first step of learning starts with the sharing of ideas.

GENERAL COMMENTS
Note: The numbers between brackets generally refers to the clauses in the new policy.

1. The Research Outputs Policy, 2015, will be implemented from 01 January 2016. The 2015 output will already be assessed on the new policy, but certain aspects of the new policy, such as the new requirements to remain on the DHET list for South African Journals, will be applied leniently to allow South African journals to adapt to the new policy. From the 2016 output onwards the full policy requirements will apply (DHET personal communication).
2. Overall, great emphasis is placed on “pre-publication” peer review (“refereeing by independent experts in the field”) as the proxy measure of quality. The Department may however in future introduce new quality measurements such as “bibliometric data, discipline specific panels of experts and post-publication peer reviews” (2.4).
3. Although all journal outputs (eg, local and international) are currently subsidised on the same level, a warning signal is sounded that in future the Department may introduce a differentiated subsidy system putting more weight, for instance, on publication in “high impact” journals, or journals listed in certain indexes (2.7).
4. The Department subsidises institutions and not individuals. The point is made that institutions should be careful not to incentivise individuals as this may act as a perverse incentive (3.1). Some institutions are providing monetary incentives to authors directly linked to DHET subsidy income, sometimes even using a differentiating system, with more rewards for Web of Science (ISI) listed journal articles. Transitioning the environment to a non-incentive rich climate will rest with the institution. This should be seen as an opportunity to level the playing field of all types of research, allowing institutions to recognise and reward scholars on the quality of the research output rather than the place of a listed publication.
5. The subsidy follows the institutional affiliation of the authors, assuming that that is where the research was “carried out”. There is no proportionality with regard to contribution, and all authors are subsidised equally (3.2). Furthermore, “Author affiliation should reflect the institution where research was conducted, supported and funded” (4.3). This clarifies the question of affiliation, which was needed.
6. Greater clarity is also provided regarding the situation when authors move to another institution (3.3), and the position of “visiting scholars, fellows and retired academics” (3.4).
7. A new requirement for institutions is to “verify, by assessing peer review reports, that all their outputs have undergone a rigorous (pre-publication) peer review process” (8.2e). This can prove very problematic, since peer review reports are regarded as confidential documents by journals and publishers in order to ensure reviewers give uninhibited reports.

JOURNALS

1. The DHET subsidises only articles published in “approved scholarly journals”, meaning journals listed in “lists of accredited journals and indices”, which will be published on or before 31 January each year. This is a major change, as the names of specific indices are now omitted in the policy, leaving the door open for the introduction of new indices from time to time. Currently the Web of Science (formerly ISI) and the International Bibliography of Social Sciences (IBSS) are the approved indices. The DHET has indicated before at the National Scholarly Editor’s Forum (NSEF) that Scopus (found at http://www.scimagojr.com), ScieloSA (http://www.scielo.org.za) and the Norwegian Register for Scientific Journals, Series and Publishers (https://dbh.nsd.uib.no/publiseringskanaler/) could be additional indices, but these are uncertain. It is therefore of great concern that the applicable indices for 2016 will only be made known in January 2016, which is too late for scholars and institutions to take advantage of listed journals. The new indices will be announced soon, although they will only be applied in 2016 to assist authors in deciding where to submit their 2016 output manuscripts (DHET personal communication).
2. In addition to the approved indices, the DHET will as before maintain a list of approved/accredited South African journals (DHET List) that are not listed on any of the approved indices. It is now made clear that it is the responsibility of the Editors-in-chief to apply to DHET for the inclusion of their journals on the DHET list. The criteria for inclusion on the DHET List (5.10) contains a few important changes, eg (5.10 c) “No more than 25% of the published articles may originate from a single institution”; (5.10 f) “No more than 33% of members of the editorial board may originate from a single institution”; and (5.11 e)” New journals must also have a publication history of at least three years.” These changes were probably introduced to prevent institutions from publishing “in house” journals with the purpose of taking advantage of the subsidy system, but the 25% limit for articles from a single institution may prove to be very difficult for many journals, particularly single faculty journals, eg veterinary science. These criteria also make it much more difficult to start new journals in emerging disciplines.
3. It is important to note that the DHET List criteria (5.10) will only apply to DHET List journals, and not to journals on approved indices (DHET personal communication).
4. Journals not meeting the DHET List criteria (5.10) will be removed from the DHET list and may only re-apply after two years. A loss of accreditation will be disastrous for the journal and the affected South African higher education institutions. It is also important for journals not listed in approved indices to adhere to the criteria in 5.10, to ensure the viability of the relevant research discipline in the South African context, as loss of accreditation will probably result in significantly less submissions from South African scholars.
5. What constitutes a “South African” journal (one which can remain/apply to be on the DHET List)? The current definition applied by the Academy of Science of South Africa (ASSAf) is that the owner of the journal (usually a higher education institution or a scholarly society) and the “workflow” of publishing must both be based in South Africa (ASSAf personal communication). This was confirmed by the DHET also indicating that DHET would like to support publishing skills and capacity building in South Africa (DHET personal communication). The implication is that only journals meeting the above definition of a South African journal will be included/remain on the DHET list. South African journals not listed in approved indices and migrating to international publishers must be aware of this.
6. A welcoming statement is that any changes to the “editor, title, frequency, ISSN or publication format (print/online)” will not impact a journal’s accreditation status (5.16). In the past, fear of losing accreditation hindered many journals from adapting to new publishing technology.
7. It is stated that the DHET supports the development of local language journals (5.18). There are very few non-English journals in South Africa, and it has to be seen if new ones will emerge. At least this provides some comfort for the ones that do exist – they will remain subsidised as long as they meet the DHET List criteria (5.10).

BOOKS AND CHAPTERS IN BOOKS

1. A much improved and expanded definition of what constitutes a “scholarly book” is provided in (6.1). There is ample provision made for single-author and multi-author books, original research and/or exposition of the available literature.
2. An important change to the criteria for book subsidies is that it must be peer reviewed prior to publication (6.2 b). This brings books in line with journals, but will have definite implications for publishers and increase the cost of publishing books, as the cost of peer review for books are much higher than for journals.
3. Additions to the types of books that will not be subsidised are: “Reference books, Dictionaries and Encyclopedias” and “Introductions and Conclusions, unless the entire book, as a unit is being submitted for subsidy claim” (6.3 c).
4. Probably the most important departure from the old policy is the increased number of subsidy units allocated for scholarly books. As indicated before, the new unit allocations will already be applied to the 2015 output (DHET personal communication). A book may now be subsidised to a maximum of 10 units, on a sliding scale based on the number of pages (6.4). This is an important development giving recognition to research in the Humanities and Social Sciences, and should restore the book to its former importance in these fields of study. It should however be noted that the page count excludes the “references, bibliography, index and appendices” (6.2e), and it also applies to the sliding scale in (6.4) (DHET personal communication). It is a pity that the policy did not rather make use of word counts, as page size and type font may have a significant effect on page numbers, and in e-books page numbers has become redundant.
5. A new feature in the policy regarding books is the establishment of an independent “Research Output Evaluation Panel (and field-specific sub-panels)” to evaluate all books and conference proceedings (6.6). In the past there has been much criticism of the DHET for the non-transparent manner in which books were evaluated. These panels have been piloted in 2014 and by all accounts it is a much improved process.
6. “Clear and unambiguous” evidence of pre-publication peer review must be provided with the book for the evaluation process (6.8). It is not clear, however, what the minimum requirements would be. Presumably a declaration by the publisher/editor in the book would be sufficient as peer-reviewer identities and reports are regarded as confidential. Publishers will probably be well advised to formulate official peer review policies for books (as they indeed have for journals). In (6.9) it is also hinted that DHET may in future “develop a list of reputable publishers”.
7. Another new requirement is “a written justification (maximum 500 words) signed by the author of the book or the general editor explaining the contribution that the book makes to scholarship.” The full requirements for the “justification” are described in 6.10 – 6.11. Presumably this can also be published in the book itself.
8. The opportunity is also created for later editions (6.12) of presumably previously subsidised books, and Dissertations and Theses (specifically excluded for subsidy in 6.3 a) to be subsidised if at least 50% new research has been added or “reworked” (6.13). Evidence must be provided for the new content/reworking.

CONFERENCE PROCEEDINGS

The dispensation for conference proceedings is also substantially amended and improved. Its model is a mixture of the journal and book models, with approved indices for conference proceedings and evaluation and approval of individual proceedings (meeting certain criteria).

1. “Approved conference lists or other indices” for conference proceedings will be published by DHET before 31 January each year (7.2). Mention is made of a consultation process to determine which indices would be approved, but this may prove to be an onerous process.
2. The criteria for subsidising a conference proceeding is much expanded (7.4). Most importantly it is now made clear that only full articles and not abstracts will be subsidised, and pre-publication peer review is now a requirement (7.4 b). Another new criterion is that more than 60% of contributions in the proceeding must be from multiple institutions (7.4 f), and that the conference must have an editorial board with a “significant majority” (undefined) of members beyond a single institution.
3. The principle that a conference proceeding article published in an accredited journal will be treated (and therefore subsidised) as a journal article, ie, one unit, is maintained (7.5), as opposed to the subsidy value of 0.5 units per article in a conference proceeding (7.6). It is clear that DHET prefers journal articles over proceeding articles, probably because journals are more structured, particularly with regard to peer review, with a better chance of maintaining a better quality.
4. The DHET will also establish a list of South African approved conference proceedings, which will be compiled from the normal submission of proceedings for subsidy purposes (7.9). Although not explicitly stated, it is assumed that this South African list will also be published before the end of January each year.
   Evidence of “clear and unambiguous…pre-publication peer review” must be provided with every application for subsidy. It is not clear in what form this must be, perhaps a policy approved by the editorial board and a declaration by the conference editor that it was adhered to?

CONCLUSION
The new policy has brought many clarifications, and introduced a few important changes, particularly with regard to the accreditation of scholarly journals and the subsidy unit value of scholarly books. DHET seems focused on rewarding quality research. The dispensation for scholarly books in particular is much improved, and it should stimulate the proliferation of research publication in books. The quality requirements for scholarly journals were made harder to attain and these will probably hamper the development of new journals, which will perhaps be an unintended consequence of these amendments.

Pierre JT de Villiers
Managing Director, AOSIS
Disclaimer: AOSIS (Pty) Ltd disclaims all liability for any loss, damage, injury or expense however caused, arising from the use of or reliance upon, in any manner, the information provided through this news item and does not warrant the truth, accuracy or completeness of the information provided.