Wrapping up

This week you built the core toolkit of evidence-based impact evaluation. You can now read a charity evaluation, trace a cost-effectiveness model, and build a rough cost-effectiveness estimate of your own—skills that most donors never develop.

But you may have noticed a limitation. Everything this week relied on measurable outcomes, strong evidence, and short causal chains. GiveWell’s approach works brilliantly for direct health interventions. It struggles with advocacy, policy change, systemic reform, and anything where the feedback loops are long and the outcomes are hard to quantify.

Next week, we’ll extend your toolkit into those harder domains. You’ll learn to reason under uncertainty using expected value, calibration, and forecasting—skills for making good decisions when the evidence is limited but the stakes are high.

Summary

This week you learned to:

• Read a charity evaluation critically, identifying the chain of reasoning and its weakest links

• Recognise where evidence sits on the hierarchy of reliability

• Trace the logic of a cost-effectiveness model and identify where judgement calls enter

• Explain how moral weights determine which interventions get funded

• Build a rough cost-effectiveness estimate and identify its key unknowns

• Engage with critiques of top-down charity and the case for focusing on comparative advantage

The evidence-based approach to doing good has directed billions of dollars to interventions that have saved hundreds of thousands of lives. Understanding how it works—and where it struggles—is essential to thinking clearly about impact.