“The PASTEUR4OA project is focused on Open Access policy developments and is undertaking a number of activities relating to policy, including mapping policies and policy-related activities, and engaging with policymakers and providing them with information about the general policy picture and what makes a policy effective.
Work Package 3 involved a set of tasks as follows:
- Describe and enumerate the policy picture in Europe and around the world
- Rebuild ROARMAP, the registry of OA policies, including the development of a new, detailed classification scheme that describes policy elements
- Collect data on the levels of Open Access material in institutional repositories around the world
- Measure policy outcomes and analyse what elements of a policy contribute to its effectiveness
The project sought out policies that exist but had not been registered in ROARMAP, and added more than 250 new entries to the database. The total number of policies globally is now 663 (March 2015), 60% of them from Europe. Of these, approximately two-thirds are institutional policies and about 10% are funder policies. Over half are mandatory, requiring some action rather than simply requesting it and over 60% of these mandatory policies are European.
ROARMAP, the policy registry, has been rebuilt with a new classification scheme for policies that records far more detail about them than before and permits much more extensive search functionality than previously. The scheme includes criteria for deposit and licensing conditions, rights holding, embargo lengths and ‘Gold’ Open Access publishing options.Links to policy documents are provided. Repository managers at policy institutions were contacted to check that we had the correct details for their policy and where necessary corrections were made. As it stands, at the end of this period of concentrated and meticulous work, ROARMAP reflects an accurate and detailed picture of the Open Access policy situation around the world.
The project also examined policy effectiveness. Three main exercises were undertaken. First, deposit rates were measured for articles in the repositories of both mandated and non-mandated institutions, and compared to the total number of articles published from these institutions. The material was identified as Metadata-Only, Full-Text, Open Access and Restricted Access. Open Access and Restricted Access are subsets of Full-Text and together comprise the whole of that category. Restricted Access means full-text articles that are showing only their metadata, with the text itselfclosed off, and are usually in this state for a period of embargo.
Across all institutions, more than three-quarters of published articles are not deposited at all, 8% are Metadata-Only, 3% Restricted Access and 12% Open Access. The rates vary by discipline. Deposit of Open Access material was over four times as high (14%) for institutions with a mandatory policy than for those without (3%). The top 20 institutions (all mandated) in terms of amount of repository content are listed. The top five are the University of Liège (Belgium), Instituto Politecnico de Bragança (Portugal), the National Institute of Oceanography (India), University of Pretoria (South Africa) and the University of Minho (Portugal).
Second, the time lag between publication and deposit of articles (deposit latency, which may be negative if the article is deposited before publication) was measured. Open Access items tend to be deposited later than Restricted Access ones, and latency periods tend to be longer in mandated institutions than in non-mandated ones (though deposits themselves are four times higher), probably because authors who deposit voluntarily are self-motivated and will do it early.
Third, we examined the deposit rate in relation to different policy criteria:
- Positive correlations were found between Open Access and Restricted Access deposit rates and the following policy criteria: Must deposit, Cannot waive deposit, Link to research evaluation, Cannot waive rights retention, Must make item Open Access
- Negative correlation was found with Cannot waive Open Access
- Significant correlation was found between Open Access deposit rate and Must deposit and Cannot waive deposit
Fourth, we examined the correlation between deposit latency (specifically, the latency of deposit within the first year after publication) and different policy criteria. There is positive correlation between early deposit and Mandate age, Cannot waive rights retention and deposit immediately. We found significant correlation between early Open Access deposits and the age of the mandate: that is, the longer a mandatory policy has been in place, the more effective it can become.
As the numbers stand at the moment (March 2015), there are not yet enough OA policies to test whether other policy conditions would further contribute to mandate effectiveness. The current findings, however, already suggest that it would be useful for future mandates to adopt these conditions so as to maximise the growth of OA.
This analysis provides a list of criteria around which we recommend policies should align:
- Must deposit (i.e. deposit is mandatory)
- Deposit cannot be waived
- Link deposit with research evaluation”