Three World Bank studies this week, in three different countries, found the same thing: the intervention wasn't touching the actual decision. Nutrition education went to women who don't control food budgets. Employment programs targeted women who can't arrange childcare. Rights were handed to refugees who can't afford to move. Development economics may have a systematic habit of funding the visible problem instead of the real one.
In each case, an intervention was designed around a visible surface condition: ignorance, culture, legal status. The actual chokepoint was control — over money, over domestic logistics, over physical mobility. The surface condition was measurable and fundable. The real constraint required either redistribution or structural change, which is neither. What didn't change across all three studies is the organizations funding the programs, the vocabulary they used to describe the problems, and the gap between what the evaluations found and what the next program will probably target.
Whether the World Bank's next funding cycle in any of these three areas changes the program design, or funds the same intervention with a new name.