Research Partner

GiveWell is an independent non-profit organisation that, since its founding in 2007, has assessed which aid programmes against extreme poverty and the diseases that come with it achieve the most per donation. GiveWell runs no aid programmes of its own: it researches, publishes the calculation behind every assessment, and makes grants itself. The organisations it knows best, and whose impact it is most confident about, it lists as top charities. Alongside that list it runs three giving funds of its own, among them the All Grants Fund, which can also finance opportunities outside the list.

Every funding opportunity is assessed against the same simple alternative: handing the same sum to very poor households as cash. How much more a programme achieves against that alternative gives a figure, and funding goes to everything above a published threshold. That threshold is not a fixed quantity but is adjusted as the available funding changes: in 2025 it stood at 8 times, and in early 2026 GiveWell lowered it to 6 times. Ahead of that calculation sits a multi-stage review: a pre-selection based on publicly available data, then access to internal monitoring data, evaluation reports and financial information, then a cost-effectiveness model built for each programme, complemented by qualitative assessments of scalability, organisational capability and sustainability. All assumptions, models and uncertainties are documented openly and updated regularly.

In its 2025 grantmaking year, which ran from February 2025 to January 2026, GiveWell approved 131 grants to 69 organisations totalling $418 million across 30 countries. Of that, $309 million went to programmes other than its top charities, and more than 200 formal grant investigations were launched in the same year. Because a threshold decides rather than a fixed list of recipients, GiveWell can also meet gaps that open at short notice: in response to cuts to US foreign aid, it made 25 grants totalling around $53 million in 2025 explicitly for the needs those cuts created. The research team behind this grew from 38 to 79 people by mid-2026.

GiveWell funds its own operations from voluntary contributions by donors and keeps that money strictly separate from the donations passed on to the organisations it has assessed. From its own funds, 100% reaches the recipient organisations after transaction fees, and GiveWell uses none of it for its own operations. The largest funder behind the recommendations is Coefficient Giving, formerly Open Philanthropy. A large share of the grant funding therefore rests on a single source. The assessment of the programmes stays separate from it, and that separation is what independence rests on: what gets funded follows from the published models, which anyone can recalculate.

The recommendations rest on randomised controlled trials and robust impact measurement. Uncertainties are analysed and named in every cost-effectiveness model; method documents, critical assumptions and potential sources of error are disclosed, and GiveWell keeps a standing page for its own mistakes. Because evaluation and grantmaking sit in one pair of hands here, verifiability rests on that publication: each grant gets a page with its reasoning and cost calculation, and every grant is listed in a public database. What can be checked is therefore not only the individual decision but the average result: across all grants made in 2025, estimated impact stood at 16 times the cash alternative, roughly double the threshold then in force. Giving What We Can assessed GiveWell as an evaluator in 2025 and concluded that its recommendations and grantmaking approach are well suited to maximising the impact of a donation in global health.

Wealth for the World makes no funding decisions of its own; it follows GiveWell's research. Each payout period, 2% of the fund's assets are paid out, whether the period ends in gain or in loss. At the end of the period, donors determine in the Impact Portal which partner receives the share attributable to their own donation. Anyone who would rather not settle on a single organisation gets our standing recommendation, the All Grants Fund: it is the broadest of GiveWell's three funds and can also finance work that appears on no list of vetted organisations, from building up a programme to getting a better drug approved to a funding gap that opens at short notice. Payouts run through Effektiv Spenden, so 100% of every payout euro reaches the partner organisations.