After two years of intensive research, Invest in Open Infrastructure's project titled “Investigating 'reasonable costs' to achieve public access to federally funded research and scientific data” has concluded. The “Reasonable Costs” project delivers crucial insights into one of the most pressing questions in scholarly communication: What does it actually cost to make federally funded research immediately and freely accessible? The findings, supported by a $299,454 National Science Foundation (US) grant, paint a complex picture of an ecosystem in transition—one where the path to compliance with public access mandates is far from straightforward.

Read the original grant proposal

The Challenge: Beyond the Nelson Memo's Vision

The 2022 White House memorandum on public access to federally funded research set an ambitious agenda: immediate free access to all federally funded research outputs, including publications, data, and code. But as IOI's research team, led by PIs Katherine Skinner and Kaitlin Thaney with team members Lauren Collister, Jennifer Kemp, Sarah Lippincott, Eric Schares, Karl Huang, Cameron Neylon, and Gail Steinhart, discovered through extensive fieldwork with 25+ academic institutions and major data repositories, the details matter when implementing this agenda.

Even when digital research outputs are free to access for users, the publication and management of these resources can be costly. The question of who should pay—and how much—for enabling publication, dissemination, and curation of research outputs remains one of the most challenging aspects of the transition to immediate public access. Furthermore, the concept of “cost” goes far beyond monetary exchange; it includes labor, communications, and workflow systems. This project sought to better understand how the costs associated with federal policy implementations in 2026 and beyond could impact particular stakeholder groups, including repositories and research institutions

Three Pillars of Investigation

The project tackled this challenge through three interconnected research streams:

1. Landscape Analysis: Defining Costs vs. Prices

The team began by establishing crucial definitions and synthesizing existing knowledge about costs and pricing in scholarly publishing and data repositories. Their work revealed a critical gap: while "cost" refers to the actual expenses incurred by publishers and repositories in providing public access, "price" represents what authors and institutions pay in the marketplace—and these figures often bear little relationship to each other.

Key findings from the landscape analysis include:

  • Cost information remains largely unknown across publishers and repositories
  • Price reporting is inconsistent and difficult to generalize
  • Disciplinary variations are enormous
  • Most publishing business models still derive from or depend largely on traditional subscription formats

For data repositories specifically, the team identified that scaling up to accommodate all federally funded research data would create significant business model challenges across different repository types, from generalist platforms to specialized subject repositories to institutional repositories.

The team published two papers delving into what was known about the cost and price of public access to scholarly publications and to research data: 

Additional findings and observations on these topics were shared in blog posts: 

Team members also presented these findings at several conferences and webinars: 

  • Skinner, K., Steinhart, G., & Collister, L. (2024, March 22). Responding to the Nelson Memo: Investigating Cost and Price in OA Data and Publishing. Year of Open Science Culminating Conference, Online. Slides. 
  • Steinhart, G., Schares, E., & Skinner, K. (2024). Navigating the future of data sharing: The impact and cost of expanded public access requirements. IASSIST & CARTO 2024, Halifax, Nova Scotia, Canada. Slides.
  • Skinner, K. & Thaney, K. (2024, January 9). "Investigating Cost and Price in OA Publishing." Federal Open Science Hour. Slides.
  • Collister, L. & Kemp, J. (2025, May 7). “‘Reasonable Costs’ for publishing: What do we know now, and what can library publishers help us find out?” Library Publishing Forum. Virtual. Recording.

2. Repository Perspectives: The Human Factor

Through surveys and interviews with eight major repositories including Dryad, Figshare, Harvard Dataverse, ICPSR, Knowledge Commons, OSF, Purdue University Research Repository, and Zenodo, the project confirmed that labor, not technology, represents the biggest cost factor in repository operations.

Repositories employ experts to do curation, programming, development, advocacy, and metadata, and repository leaders confirmed that most of these tasks can't be automated and must be done by people with specific expertise. This finding challenges common assumptions about the scalability of digital infrastructure and highlights the irreplaceable role of human expertise in maintaining quality scholarly communication systems today.

Repository leaders expressed particular concern about potential increases in demand straining not just technical systems but the essential (and expensive) human resources required for curation, preservation, and maintenance. Importantly, many repositories currently rely on business models that do not include fee structures that would increase their revenues proportionally to the number of deposited units received from an institution or researcher (e.g., “deposit fees”). Instead, many depend on tiered membership models and other contribution frameworks that are not designed to expand directly according to use. This means that as the costs associated with running a repository increase due to rising demand from federal policies, many repositories will be running at a deficit, leading to critical, system-level sustainability challenges.

These findings were presented in the following venues: 

  • Skinner, K. & Collister L. (2025, January 23). "Supporting Public Access: The Cost and Price of U.S. Institutions' Changing Research Workflows.” ICPSR webinar series. Recording.
  • Collister, L., Skinner, K., & Steinhart, G. (2025, June 16). "Repositories in the US Federal Funding Workflow: Lessons from the 'Reasonable Costs for Public Access' Project." Open Repositories 2025, Chicago, IL, USA. Slides.

3. Institutional Workflows: The Reality on the Ground

Perhaps the most revealing aspect of the project resulted from the team’s ethnographic research with 27 diverse institutions—from liberal arts colleges and R1 and R2 universities to Ivy League schools, including land-grant universities, HBCUs, and specialized research labs. The research team recruited and assembled groups of library, IT, and research office staff members in each institution and then worked with each of these groups to map their actual institutional workflows for supporting federally funded research and open sharing requirements. Once the workflow diagrams were complete, the participating institutions engaged with each other through facilitated sharing sessions in which they compared and contrasted these workflows across institutional boundaries.

The findings reveal a system that attempted to cover the needs of the institution but was frequently marked by gaps, delegation, and last-minute scrambling:

  • Institutional sponsored research workflows are highly unstandardized; their forms largely depend on local structures, relationships, and politics
  • Most institutional groups cited a belief that the compliance burden for public access is borne by individual Principal Investigators
  • "Wait-and-see" approaches dominate, with libraries appearing to be the unit most likely to take proactive steps to support public access
  • Cross-unit collaboration is happening through working groups and referrals, but institutional leadership can lack visibility into these efforts
  • Repositories enter workflows late in the process, typically at the publication phase
  • Data management plans are frequently copied and pasted or contain only bare-minimum responses, creating stress for everyone involved

The team published two papers and a supplemental dataset of visualizations of institutional workflows: 

  • Steinhart, G., Collister, L., Kemp, J., Lippincott, S., & Skinner, K. (2025). Between the Memo and the Mandate: Institutional Perspectives on Public Access Readiness. https://doi.org/10.5281/zenodo.15706351
  • Collister, L., Lippincott, S., Kemp, J., Skinner, K., & Steinhart, G. (2025). Preparing for Public Access: Mapping Institutional Workflows for Sponsored Research Success. https://doi.org/10.5281/zenodo.15611424

Additional synthesis was shared in a blog post: 

These findings were also presented in the following outlets: 

  • Collister, L., Dollar, D., Rachlin, D., & Walsh, M. (2025, April 4). “Realizing public access to federally funded research: Views from a range of academic libraries.” (Panel.) Association of College and Research Libraries, Minneapolis, MN, USA. Slides.
  • Skinner, K., Collister, L., Lippincott, S., Kemp, J., & Schares, E. (2025, April 15). “Investigating Price, Cost, and Workflows Associated with Public Access” Federal Open Science Hour. Slides.

An article on this section of the project has been submitted to an academic journal and is awaiting review.

Practical Tools: Making the Invisible Visible

The project didn't just identify problems — it created practical solutions. Recognizing that predicting the Nelson Memo's impact on activities and bottom lines was one of the most perplexing challenges for publishers, repositories, researchers, and research institutions alike, two complementary tools were specifically designed for research institutions.

The Publishing Profiler: Visualizing Research Output Footprints

The Publishing Profiler represents a tool to help institutions understand their federal research publication landscape. Built by consultant Eric Schares and expanding on his earlier work at Iowa State University, this interactive dashboard combines data from Dimensions and OpenAlex to provide comprehensive insights into institutional publishing patterns.

Using an eight-year analysis period (2016-2023) to capture trends and context, the Publishing Profiler reveals:

  • Total publication output with breakdowns by corresponding authorship status
  • Federal funding acknowledgments extracted from acknowledgment text and metadata
  • Nelson Memo-implicated publications—the crucial intersection of corresponding authorship and federal funding that would fall under new policies
  • Publisher and journal patterns showing where researchers are publishing and how this is changing over time
  • Open access status across different forms of OA
  • Funding agency breakdowns allowing deeper dives into specific federal funders

For example, Iowa State University's 2023 data showed 2,776 total published articles, with 48% having ISU corresponding authors and 55% of those acknowledging federal funding, resulting in 733 articles that would be subject to the Nelson-era open-access policies.

The tool's practical applications are immediate: libraries can gauge future demand for institutional repositories based on the percentage of non-OA publications that will need compliance support, while research offices can help investigators budget for article processing charges or plan to publish in fully open access journals.

The Cost Calculator: Understanding Real Institutional Expenditures

The Cost Calculator, developed in partnership with the Curtin Open Knowledge Initiative (COKI), is a prototype framework to help institutions estimate their actual expenditures across four categories of open access costs:

  • Output-level expenditures: APCs, page charges, and other publication fees tied to individual research outputs
  • Publisher payments: Costs for "Read and Publish" agreements and other transformative publishing deals
  • Contributions to services/systems: Support for platforms and repositories where payments aren't tied to specific institutional outputs
  • Internal expenditures: Institutional repository systems, platforms, and staff costs for managing agreements and deposits

Released as a transparent, shareable spreadsheet framework, the Cost Calculator allows institutions to blend their internal data with public information to create comprehensive cost estimates. However, during the development of this tool, we realized that its utility is limited by institutional data management practices.

The Data Challenge: Why Better Tools Need Better Data

The Cost Calculator's pilot testing with institutions uncovered significant systemic issues that go beyond any single tool's capacity to solve. Institutions across our Reasonable Costs study consistently reported:

  • High complexity in internal systems: Each institution handles data differently, with frequent internal inconsistencies
  • Missing article-level data: Both public and internal datasets lack crucial information for accurate calculations
  • Tracking challenges: Manual, messy processes are divided across multiple staff and systems
  • Agreement complexity: Read and Publish deals make it nearly impossible to separate the exact costs for publishing components
  • Overestimation risks: Multiple institutional affiliations on publications and authors create double-counting problems

The project team's conclusion was stark but important: As long as institutional data about publishing costs is unstandardized and spread out across staff and systems, it will likely be too costly for an individual institution to use any tool to better understand its spending patterns. Instead of a 'cost calculator,' institutions need strong, standardized models for their accounting and reporting around research output publishing and deposit practices.

Key Recommendations: A Path Forward

The research yielded several actionable recommendations for institutions preparing for the new landscape:

Empower librarians with knowledge and support: Including librarians and IT staff earlier in the proposal process helps researchers manage and share data more effectively throughout the project lifecycle. For example, one small institution in the study sends its libraries a list of PIs and awards during the proposal phase—a simple change that dramatically improves collaboration.

Institutional cooperation is difficult but powerful: Cross-unit collaboration is happening, but university leadership often lacks visibility into these efforts. When research offices discover the extent of librarians' work supporting public access, they want to strengthen these partnerships.

Last-minute data management plans don't work: Copy-pasted DMPs and neglected practices create unnecessary stress and complications. Templates, repository guidance, and engaged colleagues can help, but research teams need to value and promote data sharing from the project's beginning.

Additional Critical Observations: The Broader Infrastructure Challenge

The project revealed several systemic issues that extend beyond individual institutional challenges:

The metadata crisis: Poor and incomplete metadata emerges as perhaps the single most important factor keeping the scholarly communication field in the dark about costs, prices, and usage patterns. This is particularly acute for funding acknowledgments, where the field relies on inconsistent acknowledgment text rather than standardized metadata across all output types.

Repository sustainability at risk: Repositories operating on membership models or contribution frameworks that don't scale with usage face significant danger. They may experience steep increases in demand while maintaining flat revenues that can't offset increased operational costs.

The US federal funding reality: Unless funders' research budgets grow, allowable costs will impact the amount of funding available for direct support for research. In 2025's dramatically reduced federal funding environment, this risk has become acute.

Transparency challenges: Quantitative information on both cost and price remains difficult to find, with few incentives for publishers and repositories to make this information public. The team suggests that developing "reasonable cost" policies analogous to federal travel restrictions would require much clearer evidence and documentation of publication and deposit costs and their variance.

Labor as a key factor: Human expertise for data curation, repository management, and scholarly communication support cannot be easily or quickly scaled up. As demand increases, this creates fundamental bottlenecks that technology alone cannot solve. We must invest in the people who do this work and who teach others, recognizing the value of human contributions in a funding and research landscape that can sometimes prioritize tool and technology development. 

Looking Ahead: Questions That Remain

While the Reasonable Costs project has provided crucial clarity on many aspects of public access implementation, it has also highlighted areas requiring continued attention:

  • How can institutions better integrate public access planning into their research workflows from day one?
  • What sustainable funding models will emerge for the human expertise essential to repository operations?
  • How can cross-institutional collaboration help smaller institutions meet compliance requirements?
  • What role should federal agencies play in supporting the infrastructure needed for immediate public access?

As the research community continues to navigate the transition to immediate public access, the Reasonable Costs project provides both a roadmap and a reality check. The path forward requires not just policy changes but fundamental shifts in how institutions organize themselves around the values of open science. Although the project wrapped up with a final public webinar on September 4, 2025, the IOI team envisions that the learnings and methods will be useful for many years to come. If you’re interested in extending the use of the project materials or learning more about the project itself, please contact us at research@investinopen.org. 

The complete collection of publications, presentations, and tools from the Reasonable Costs project is available at investinopen.org/data-room/reasonable-costs/. The project was supported by National Science Foundation Grant No. 2330827.

Posted by Lauren Collister