Philanthropic Funding Methodology

Data sources

To develop a comprehensive database of philanthropic funding for the ocean during the time period 2010-2020, Our Shared Seas (OSS) gathered grant-level funding data from four sources:

  1. Direct outreach to foundation staff of the top 60 marine funders
  2. Candid, a 501c3 nonprofit organization that tracks philanthropic funding from the longtail of foundations to grantee recipients
  3. The European Foundation Center, which provided estimates of European funders in the marine sector by issue area and geography
  4. IRS Form 990 tax documents and foundation websites


Several steps were completed to ensure that data were comprehensive, non-duplicative, and appropriately categorized by geography and topic. New data was acquired for the longtail of 2010-2014 grants and all 2017-2020 grants to supplement the existing OSS ocean funding database. Whereas the baseline database contained information with a heavy focus on U.S.-based foundations, the 2021 dataset added several new, non-U.S. foundations through direct outreach to known marine funders headquartered outside the United States.

The OSS ocean funding database, which includes the “longtail” of ocean funding as collected by Candid, tracks fewer funders for the period 2010-2014 as compared to 2015-2018. Additionally, full records for the longtail of funding for 2019 and 2020 are not yet available due to a time lag in reporting. To better analyze trends in ocean funding over the last decade, OSS forecasted a proportion of the longtail from 2015-2018 for the incomplete records (2019 and 2020) and back-casted for the period 2010-2014 with a discount rate of 6 percent annual growth.

For consistency with the OSS existing database and to use the highest quality data, grantmaking commitments are used. For commitments larger than USD 10 million, grants were annualized over the life span of the commitment (i.e., a USD 30 million commitment lasting five years was broken into five grants of USD 6 million per year).

Because data were compiled from multiple data sources, the study team identified and removed duplicates through both automated and manual processes. During duplication removal, the most specific and comprehensive records were maintained. Where appropriate, grants obtained through direct outreach to foundation staff were retained, replacing grants received from Candid or other sources.

Grants were categorized initially using a keyword search and then subsequently reviewed manually by the study team using expert judgement with regards to geography and topic. The aggregate number of Candid grants with commitment amounts of USD 10,000 or less were consolidated into a single entry and were not categorized at the grant level. In cases where grants pertained to multiple geographies or topics, reviewers divided the grant amounts proportionally. Where there was insufficient detail to categorize a given grant, an “unspecified” category was used for one or both taxonomies. During this manual review of data, additional duplicate grants were removed by reviewers. In 2021, intermediaries such as regrantors and fiscal sponsors were identified and extra effort was taken to avoid double counting of such grants.

After a manual review of grants, a final validation review was completed. Gaps in funding were identified and addressed based on the study team’s understanding of the funding landscape. The study team’s understanding of foundations’ and recipients’ program areas was additionally used to validate categorization of grants; geographies and topics were standardized across program areas and years to remain consistent.

It should be noted that there are certain elements not included in this analysis, including government funding and comprehensive accounting of individual giving, including through Donor Advised Funds (DAFs). While the Our Shared Seas team expanded global representation in the current edition of the report, coverage of foundations based in Asia is likely incomplete and thus an underestimate.

Considering these refinements, readers should reference the current report for comprehensive 2010-2020 funding figures. Comparing results across previous editions of this report (i.e., version one in 2017 and version two in 2019) is not an apples-to-apples comparison. However, the current report edition includes a consistent methodology for estimating funding across the full time period of 2010-2020.