Where the money goes
Funding insights at a level no single funder can see from inside its own portfolio.
Grant funding flows are one signal of how the research ecosystem values and takes responsibility for the infrastructure it shares.
The data and analyses we published in 2024 and 2025 primarily focused on better understanding the amount, impact, and limitations of grant funding to Open Infrastructures (OIs). In 2026, we are looking more closely at a related factor: how much funding is supporting infrastructure elements directly versus how much funding is supporting research that leans on and invokes infrastructure, but doesn’t contribute to it financially.
This section maps the funding dependencies for infrastructure and how narrowly the sustaining funding is sourced. We hope our analysis offers funders, institutions, and infrastructure providers a shared, independent picture of patterns that none of them created alone, and that none can address alone.
Our full data collection and preparation methods are described in detail in the documentation that accompanies the raw data — this is available for download.1 We encourage readers to explore the full documentation for additional information on how we processed the data, as well as key assumptions and their likely trade offs. The funding described in this report is limited to grants we were able to identify through public sources and does not include other sources of revenue or funding for these infrastructures.
All financial data in the report is quoted in US dollars.
Shares are of $663M across 965 dollar-bearing award clusters, out of 1,065 clusters in total. The remaining 100 carry no allocatable dollar figure and are counted by presence only.
Grant funding flows are one signal of how the research ecosystem values and takes responsibility for the infrastructure it shares.
Most grant funding supports research that depends on open infrastructure; far less sustains the infrastructure itself.
We classify grants as direct (funding an OI), adjacent (funding research that depends on an OI), or adoption (funding an entity to deploy an OI). The imbalance between direct and adjacent is the clearest signal in the data of how funders allocate funding: more heavily toward the research, and lightly toward the infrastructure beneath it. The ratio varies widely by infrastructure; for some, a few hundred thousand in direct support underwrites tens of millions in dependent research.
$325K direct
$39K direct
$1.6M direct
$724K direct
$371K direct
$297K direct
Across the ecosystem, every $1 of direct funding that reaches an infrastructure sits beneath $1.65 of funded research that depends on that infrastructure. This could be interpreted in two ways:
Modest infrastructure investment carries a much larger research enterprise
The wider the gap, the more output rests on support that hasn’t grown with it
Within that thin slice of direct funding, most pays to build, not to run.
Of the $192M reaching infrastructure directly, the majority funds new development rather than sustained operation. This is not a failure of any funder’s judgement — grant proposals are designed around projects, and projects tend to build things. But project-based funding leaves ongoing operations (the part that keeps infrastructure available year after year) structurally harder to fund.
Research relies on open infrastructure globally. The funding that sustains infrastructure tends to come from single sources.
Direct funding is concentrated by geography: most infrastructures draw their sustaining grants from one country even when they support global research. Seen two ways — by geography, and by funder — the pattern is the same, and it is not one any single organisation created or can resolve alone.
Each bar counts the infrastructures a funder has supported directly, of the 52 that received any direct funding.
Infrastructures funded directly from a single country
Of the 34 infrastructures with more than one direct grant, 22 receive that support from one country, while the research depending on them is funded from 2.6 countries on average. Single-source funding ties an infrastructure’s continuity to one country’s budget cycles and policy priorities, however global its users.
Distinct infrastructures the National Science Foundation (US) alone has funded directly
A small number of agencies and funders underwrite much of the ecosystem: NSF appears in the direct funding of 13 different infrastructures, IMLS 10, Mellon, Sloan and EC 8 each. When one funder’s priorities shift, the effects reach many infrastructures at once, and the risk is correlated.
The asymmetry up close: Award counts by country.
Nine countries have funded research that depends on arXiv; one provided grant funds for arXiv itself. For Open Science Framework (OSF), this ratio is seven to one. The reliance is distributed across the research world; the funding that sustains the infrastructure is not.
Where that concentration meets current risk.
Since January 2025, US federal research funding has faced sustained pressure: agency budget threats, indirect cost caps, and the near-elimination of IMLS. $42.1M in active federal grants run to open infrastructure through this period. The point is bigger than the US: infrastructure the world depends on, sustained mainly by one government, is only as durable as that government’s priorities.
| Infrastructure | Federal funder(s) | Active grant value |
|---|---|---|
| ICPSR | NSF · HHS | $18.1M |
| arXiv | NASA · NSF | $6.4M |
| Dataverse | NIH · HHS | $3.3M |
| Zenodo | HHS | $2.3M |
| Dryad | HHS | $2.2M |
| Galaxy | NIH · HHS | $1.9M |
| OSF | NSF | $1.5M |
| + PUDL, DMPTool | $0.6M | |
| Multi-recipient grants span several OIs | $5.8M | |
| Total at risk | $42.1M |
The grant awards dataset used for the analysis presented here is available online at https://doi.org/10.5281/zenodo.21401121, together with full methods documentation and a data dictionary. The code used for processing and analysis are available online at https://github.com/investinopen/state_of_open_funding_data_processing/.
Interactive dashboards allowing direct exploration of the data are available at https://investinopen.org/grant-funding-data-dashboard/.
Questions or concerns? Please contact us at research@investinopen.org.
Feedback