The best solutions are far more effective than others

Abstract

This article examines the potential for significant variation in the effectiveness of solutions to social problems. The article argues that the best solutions within a given area often achieve considerably greater impact per unit of effort compared to the average solution. Several studies are reviewed that demonstrate this pattern across a variety of fields, including global health, education, and climate change. The article explores reasons for this variation, including the lack of strong feedback loops between impact and resource allocation, the potential for regression to the mean in effectiveness estimates, and the possibility that the best solutions are simply not well-measured or documented. The article advocates for a “hits-based” approach to finding solutions, prioritizing those with high potential upside even if they carry a higher risk of failure. Several frameworks and methodologies are discussed for identifying promising solutions, including upside/downside analysis, importance-neglectedness-tractability analysis, and bottleneck analysis. – AI-generated abstract