Microfinance's Broken Promise: How RCTs Rewrote Development Orthodoxy
When randomized trials finally tested development's favorite intervention, the results demanded a complete rethinking of what credit can achieve.
External Validity: Can Results from Rural Kenya Predict Outcomes in Urban India?
Why promising development evidence often fails to travel, and how to predict when it will
The Hawthorne Effect in Development: When Observation Changes Everything
Rigorous evaluation may systematically inflate results that disappear when programs scale beyond intensive observation.
Randomization Failures: When Your Control Group Isn't Really a Control
How contamination, attrition, and selection problems silently invalidate your experimental estimates
Regression Discontinuity: Sharp Evidence from Arbitrary Cutoffs
Arbitrary eligibility thresholds transform administrative rules into natural experiments, delivering rigorous causal evidence when randomization proves impossible.
Spillover Effects: The Hidden Force Distorting Your Treatment Estimates
Why your control group may already be treated—and how this hidden contamination systematically underestimates program effectiveness across development interventions.
Why Most Development Programs Fail Before Implementation Even Begins
The hidden design failures that doom development interventions before beneficiaries are ever reached—and the institutional pressures that make them inevitable.
Power Calculations: The Math That Separates Credible Evidence from Noise
Master the statistical foundations that determine whether your impact evaluation produces actionable evidence or expensive, misleading noise.
Pre-Analysis Plans: Credibility Insurance or Bureaucratic Burden?
Why committing to analytical decisions before seeing results transforms development research credibility without sacrificing scientific discovery.
The Take-Up Problem: Why Eligible Beneficiaries Don't Participate
Understanding why eligible beneficiaries don't enroll—and how evidence-based design can close the participation gap in development programs.
Cost-Effectiveness Analysis: The Metric Development Needs But Rarely Uses
Development organizations commission rigorous impact evaluations then ignore the cost data that determines whether interventions actually deserve funding over alternatives.
The General Equilibrium Problem: Why Scaling Changes Everything
Why rigorous pilot results can mislead when interventions grow large enough to reshape the markets they operate within.