The key problem this idea examines is the difficulty in making optimal resource allocation decisions, particularly in fields like global health, development, and research, where billions of dollars are spent annually with imperfect information about what will prove most impactful. The challenge lies in learning from past allocation errors to improve future decisions, potentially saving lives and resources through better-informed choices.

The Retrospective Analysis Approach

The core of the idea involves conducting a comprehensive 20-year review (2000-2020) of spending patterns against current effectiveness knowledge. This could be broken down into:

1. Mapping major spending categories and interventions
2. Applying modern cost-benefit frameworks retroactively
3. Modeling alternative allocation scenarios
4. Quantifying how much better decisions could have been with current knowledge

The goal would be to produce both concrete findings (e.g., "X dollars allocated differently could have saved Y lives") and general frameworks for future decision-making. For funders and policymakers, this could mean identifying persistent biases in allocation patterns or highlighting overlooked but highly effective interventions.

Implementation Strategy

One way to structure this initiative would be through phased implementation:

• Initial phase (6 months): Select 2-3 high-impact domains for case studies (e.g., malaria prevention, education reform), gathering historical spending data and current effectiveness metrics.
• Expansion phase (12 months): Develop broader frameworks and create tools for scenario testing across more domains.
• Long-term: Establish systems for regular updates and policy engagement as new data emerges.

A minimal viable product could focus on analyzing one specific domain with clear before/after effectiveness comparisons, establishing proof of concept before expanding.

Relationship to Existing Work

While organizations like GiveWell evaluate current charity effectiveness and the Copenhagen Consensus ranks solutions by cost-benefit, this approach would differ by adding a temporal dimension. By examining how evaluation quality has changed over time, it could reveal patterns in decision-making that existing forward-looking analyses might miss. World Bank-style retrospective evaluations tend to be narrow in scope, whereas this would attempt to create a unified, cross-domain framework.

The project's value would come from combining rigorous historical analysis with current effectiveness data to create practical decision-making tools—turning hindsight into foresight for better resource allocation.