Editor’s Note: Today’s post is by Ramazan Turgut. Ramazan is Managing Editor at the Directory of Open Access Journals (DOAJ) and serves as Ambassador for Crossref, the Directory of Open Access Books (DOAB), and OPERAS in Türkiye
During a panel discussion on open access, a volunteer editor interrupted me to express his frustration: “I am a simple man. Tell me concretely what you want me to do, and I will try to do it. But please, I cannot track all these declarations, action plans, frameworks, and principles. They make my job harder, not easier.” The room became silent. After forty minutes of debating the philosophical merits of the latest Diamond open access (OA) manifesto, no one could provide a concrete answer to his question.
That exchange occurred two years ago, and I have reflected on it frequently. In the intervening period, the proliferation of declarations has continued, but editors’ willingness to engage with them has diminished.
The chronology is now familiar to anyone working in scholarly communications. The Budapest Open Access Initiative in 2002. Berlin and Bethesda in 2003. Plan S arrived in 2018, followed by the UNESCO Recommendation on Open Science in 2021. Then came the Action Plan for Diamond Open Access and the Coalition for Advancing Research Assessment in 2022. In December 2024, the second Global Summit on Diamond Open Access produced the Toluca-Cape Town Declaration; a third summit followed in Bengaluru in February 2026. In November 2025 came the Stockholm Declarationon the reformation of science publishing. Well-meaning people signed each document. Each expressed noble intentions. But the question remains: what, precisely, changed on the ground because of any single one of them?
The uncomfortable answer appears to be very little, and the critique is now coming from inside the movement. In June 2026, Csaba Szabo published “Why the Stockholm Declaration will never work” in Royal Society Open Science, arguing that the declaration misdiagnoses the disease: the crisis stems from systemic incentives that reward volume and prestige over validity, not from profit motives alone. He is right about the diagnosis. But the same objection applies to his own conclusion, and to nearly every document on the list above. None of them carries binding commitments, a funding structure, or an audit process. They add voices to a chorus that has been singing for over two decades without producing sheet music.
This phenomenon can be described as declaration inflation. It is not the result of ill intent, but rather a structural issue. Periodically, new coalitions identify gaps in the existing framework and respond by drafting additional documents. However, because these declarations originate from diverse constituencies with varying priorities, they seldom coalesce into a unified, actionable agenda. Instead, they accumulate over time, each distinct and visible, yet collectively insufficient to effect substantive change.
The Participation Illusion
A second, more nuanced issue exists. Signing declarations can create the impression of meaningful action. For example, a journal editor who endorses a new statement may believe she has contributed significantly to open access, while a university librarian who signs a petition may feel he has advanced the cause. However, this perception constitutes a participation illusion.
Jerry Harvey’s Abilene paradox describes groups that collectively decide on a course of action that no individual actually wants, because each member believes the others support it. In the context of Diamond OA, we see this when institutions continue to pour money into subscription agreements while simultaneously signing declarations that call for a full transition to open access. Nobody is lying. Everyone genuinely supports the values expressed in the documents. But the collective behavior of the group, the actual spending, the actual hiring, the actual infrastructure investments, does not align with the collective rhetoric.
Solomon Asch’s conformity experiments in the 1950s demonstrated that individuals may publicly agree with incorrect answers if enough peers express the same view. In many parts of the scholarly publishing community, the prevailing consensus is that “Diamond OA is the future.” Articulating this position incurs no cost and demands neither institutional transformation nor budgetary or structural changes. The community has prioritized the creation of a declaration machine over a publishing machine: the former generates statements of intent, while the latter produces journals.
A colleague at a European library consortium put it to me more bluntly. “We have signed four declarations in five years,” she said. “Our open access budget has remained flat. We have added zero new staff to support non-commercial publishing. Tell me what changed.”
What the Data Actually Shows
The gap between declaration and outcome is not anecdotal. In May 2026, the US Government Accountability Office reported that federal publishing charges could cost the government up to $1 billion a year, and that annual costs may triple if historical patterns in fee increases continue. Of the nine agencies reviewed, only the National Institutes of Health had planned for these costs at all. Four years after the 2022 Office of Science and Technology Policy memo, the policy exists, and the budget analysis does not. SPARC’s response was not another declaration. It was a set of data-backed demands: distinguish public access from pay-to-publish, recognize the integrity risks of article processing charge (APC)-driven models, and invest in alternatives.
Elsewhere, we see more promising patterns. SciPost, a non-profit open access publishing platform, runs on roughly €23,000 per month, close to €300,000 a year, with a handful of staff and a published per-article cost of around €500. Its accounts are public down to the country level. It did not achieve this by signing declarations. It achieved it by building a declaration machine structurally integrated with a publishing machine: an operational framework with budgets, staff, and accountable decision-making bodies. Its persistent problem is not theory. In 2023, its expenditures exceeded its income by about nine percent, because institutions that endorse Diamond OA in principle do not fund it in practice.
Similarly, national funding models in the Netherlands (Open Science NL), France (the national plan for open science), and Latin America (SciELO, Redalyc) have demonstrated that Diamond OA can scale when there is consistent, direct public investment tied to specific performance metrics. Türkiye’s DergiPark, operated by TÜBİTAK ULAKBİM, the national academic network and information center of Türkiye’s scientific research council, hosts around 3,000 journals and close to 800,000 articles as a fully public-funded Diamond OA platform. Its metadata is now one of the core ingestion sources for OpenAlex, alongside Crossref, DataCite, and PubMed. These are not declarations. They are budget lines, technical pipelines, and editorial policies working in concert.
This contrast is instructive: declaration machines generate visibility, whereas publishing machines produce journals. At present, the Diamond OA community excels in creating visibility but faces challenges in sustaining robust publishing operations.
Five Concrete Steps
If it is acknowledged that declarations alone are insufficient, the question becomes: what should take their place? The following five shifts are proposed to transition from rhetoric to structural change.
1. From Endorsement to Binding Commitment
The Directory of Open Access Journals (DOAJ) has taken a step in this direction by revising its approach to editorial endogeny, where editorial board and reviewer pools overlap heavily with author lists. Instead of a vague statement about supporting editorial integrity, the criterion is a number: the proportion of research papers with at least one author who is an editor, board member, or reviewer must not exceed 25 percent across either of the two most recent issues. You can check it. A journal either meets it or does not. But this remains a voluntary system. What if the next major Diamond OA agreement required signatories to redirect a specific percentage of their publishing budgets to non-commercial platforms within a fixed timeline? What if signing came with an audit requirement?
David Crotty, writing in the Scholarly Kitchen in May 2026, examined why funder efforts to cap APCs have largely failedto curb publisher behavior, noting that more than 30 percent of Wellcome-funded open access papers still appeared in hybrid journals despite an explicit ban on hybrid spending. His conclusion aligns with this diagnosis: rules without structural reform of the publishing incentive system amount to treating symptoms rather than causes.
2. Scale What Already Works
The Dutch open science program, France’s national open science plan, SciELO, Redalyc, and DergiPark illustrate that sustained public funding can support non-commercial publishing at scale. These initiatives are not pilot projects; they represent long-term investments characterized by recurrent budgets, transparent governance, and documented outcomes. Rather than reinventing this model, other regions should focus on adapting and adequately funding it.
3. Shared Infrastructure, Publicly Funded
OPERAS, the European research infrastructure for open scholarly communication in the social sciences and humanities, currently constituted as a Belgian non-profit association while it pursues formal European research infrastructure status, provides a model for how a distributed network of scholarly communication services can be sustained through pooled public investment. It is not a declaration. It is a legal entity with a budget, a staff, and a reporting structure.
4. Transparent Governance
Too many initiatives in our field rely on self-appointed steering committees with unclear accountability. If a coalition is going to speak on behalf of the global Diamond OA community, its members should be elected, its finances published, and its decisions subject to public review. SciPost publishes its accounts. Most coalitions that lecture publishers about transparency do not.
5. Stop Confusing Activity with Progress
A new declaration often results in press releases, panel discussions, and increased social media engagement. These constitute activities rather than substantive progress. Genuine progress is reflected when researchers at underfunded institutions can publish without incurring article processing charges, and when editorial boards can sustain their journals without perpetual reliance on unpaid volunteer labor.
What Comes Next
It is important to clarify that declarations are not without value. The Budapest Open Access Initiative, for example, was transformative because it unified a previously fragmented movement. The issue is not the existence of declarations, but rather the tendency to regard them as the culmination of efforts instead of the starting point.
After two decades, the necessary components are clear: sustainable funding, transparent governance, shared infrastructure, and binding accountability. The central question is whether there is sufficient institutional will to implement these changes, or whether the community will continue to produce documents outlining aspirations for others to fulfill.
Drawing from my experience with DOAJ, I have observed that the most transformative moments rarely stem from the announcement of new policies. Instead, real change happens when editors revise workflows, funders adjust calls for proposals, or librarians reallocate budget lines. These incremental, unheralded decisions may not attract publicity, but they fundamentally expand what is achievable. I invite fellow editors and publishing professionals to reflect: what is one concrete step your journal took this year that mattered more than any declaration?
Therefore, my response to that volunteer editor is as follows: rather than awaiting the next declaration, focus on constructing the publishing infrastructure that renders such declarations unnecessary. Implement policy changes first, and then communicate the resulting outcomes.
Author’s note: I used an AI assistant (Grammarly) to copy-edit this post. All arguments, judgments, and conclusions are my own.
