Skip to main content Link Menu Expand (external link) Document Search Copy Copied
FOSS Foundations Directory

Some nonprofits publish their official budget for review; many also publish an annual report that gives a high-line estimate of overall expenses and revenues. Building a simple and “good enough” model here will allow inclusion of estimated finances for a lot more organizations (especially ones outside the US).

Budget Modeling

Wouldn’t it be great to have a database of revenue, expenses, and the like across the FOSS ecosystem? What are the minimal figures that we need to track - and can actually compile for a broad enough array of organizations? Work in progress - contributions wanted!

  • organization: either a foundation (legal entity), or a project (unofficial entity that uses a fiscal host)
  • year: calendar year for comparison
  • end date: month/year that a budget ends (commonly used in reporting)
  • source: where the figures were sourced from: organization, or elsewhere
  • sourceurl: specific document figures were taken from
  • updated: date the figures were copied from the source (in case someone restates an annual report, etc.)
  • total revenue
    • What general revenue categories can we commonly find in existing annual reports?
  • total expenses
    • As above: what are the top categories we can usefully track here?
  • Budgets should be stored per-entity, but should have an easy way to store figures across multiple years

Roadmap / Contributions Wanted

Budgets may not necessarily reflect actual finances, so we need to be sure to mark all data with both a source, as well as some kind of quality/range information. For example, many non-US foundations do not have public tax filings, so we need to use any published budgets or annual reports as a proxy. Many foundations produce annual reports with high-level revenue and expense numbers, but they are only estimates, not actual numbers.

  • Build the budget model; reflect yearly budgets vs. actual finances vs. annual report estimates.
  • Simplistic scrapers to detect likely “annual report” or “budget” pages.
  • Data gathering and linting.
  • Building comparisons between estimated budgets and actual 990 finances.