Over the past 18 months, university library consortia in Germany, Sweden and California have cancelled their subscriptions to Elsevier journals. This is because we have not yet reached agreement on how to support their objectives, which include making researchers¡¯ published articles immediately open access.

Elsevier deeply regrets this situation. We fully support open access. And we recognise the benefits and importance of open access to many research communities. By sharing our perspective, we hope to build support to move things forward globally.

Journal publishers serve the research community. Our role is to filter, enhance, disseminate, register and preserve new knowledge. We put significant resources into soliciting, vetting, curating and disseminating high-quality content, and then technologically enriching it to produce fully searchable, hyperlinked, easily manipulated, feature-rich articles. On average, each Elsevier article contains more than 400 embedded links for navigation, verification and citation, for example. Numerous mechanisms, from editorial independence to anti-plagiarism technologies, ensure that published research upholds the highest standards of integrity.

About 2.5 million articles are published globally each year. Roughly 85 per cent are published on a subscription basis, although preprints can be made freely available immediately and accepted manuscripts are often posted openly online after an embargo (known as green open access). The remaining 15 per cent are published on an immediate open access basis in exchange for a fee. Proponents of this ¡°gold¡± model, which has been around for 15 years, are understandably frustrated that its uptake has been slow. The reasons for this are complex, but are structural, not ideological.

³ÉÈËVRÊÓƵ

Last year, Elsevier published 470,000 articles. Only 34,000 were gold open access, but all of them could have been if authors had chosen that option. Individual articles can be made gold open access whether published in established subscription journals or in one of our new fully gold journals (we¡¯re launching one, on average, every three working days). Across both models, and on a per-article basis, Elsevier charges around or below the industry average, while our quality is above the industry average.

So why is the uptake not higher? Apart from author demand, another key factor is the common preference of large funders and institutions for green open access over gold. That is because a universal switch to gold would see research-intensive institutions and countries paying a larger portion of publishing costs because they are the ones producing the most high-quality research. Institutions with no published output would pay nothing.

³ÉÈËVRÊÓƵ

Consortia in Germany, Sweden and California want all their researchers¡¯ output to be gold open access, which we fully support. At the same time, they also want their researchers to keep reading the rest of the world¡¯s articles that, today, are still published under the subscription model. Finally, they want this package to cost no more than they have paid historically for subscriptions alone.

Specifically, California Digital Library (CDL), which represents the University of California (UC), each year pays about $11 million (?9 million) to read Elsevier¡¯s 435,000 newly published subscription articles. Of these, 1 per cent are authored by UC researchers. To publish them gold open access at market rates would cost approximately $15 million.

If the world became gold open access overnight, there would be nothing to subscribe to, and CDL would only need to pay to publish UC¡¯s articles. But this is far from today¡¯s reality. To make progress, we need to work on a solution that addresses the fact that the large majority of the world¡¯s published research is not yet gold open access.

Aiming to make such progress with UC, Elsevier offered to hold subscription prices in line with inflation and to fully fund a fivefold increase in the volume of UC-authored articles published gold open access.

³ÉÈËVRÊÓƵ

Our offer did not deliver everything that CDL was requesting, but would have enabled CDL to increase its gold open access uptake with Elsevier by more in one year than has been achieved over the past 15 years, and without incurring additional costs. We are willing to share the financial burden of a transition in a sustainable way. Unfortunately, our efforts were unsuccessful and CDL cancelled its subscription?¨C?leaving UC researchers without seamless access to newly published articles.

We believe?that the most fruitful way to achieve open access is to work constructively, step by step, country by country, based on the circumstances. There is no silver bullet. The challenges are complex, but they can be overcome with creativity, flexibility, commitment and pragmatism. In Poland, Hungary, Norway and France, we have recently concluded agreements that have both subscription and open access components. We are optimistic that we will also find ways forward with California, Germany and Sweden, and remain strongly committed to helping them achieve their objectives.

Solving such challenges paves the way to unlock further value for the research community. Huge inefficiencies exist in the $500 billion that is spent annually on academic and government research, related to grant administration, talent management, data and image manipulation, and impact measurement. The term ¡°open science¡± encompasses information-based solutions, such as research data management, facilitating collaboration, improving research integrity and evolving systems of evaluation.

While the potential benefits of open access are big, the benefits of open science are even bigger. Each percentage point of efficiency gain is worth $5 billion to the sector. More importantly, it will further advance science. If all of us in the research ecosystem engage and collaborate constructively, together we can deliver the benefits of both open access and open science.

³ÉÈËVRÊÓƵ

Gemma Hersh is Elsevier¡¯s senior vice president, global research solutions. She will be interviewed on this topic by ³ÉÈËVRÊÓƵ editor John Gill at THE¡¯s World Academic Summit in Zurich on 12 September.