On the other hand, there are large returns from improving the productivity of human-capital investments. In other words, we should try to make learning more productive. This is like shifting up the whole marginal-benefits curve, which would result in a first-order increase in income. But how can we do this? Some ideas:

* Control tropical disease. My earlier post on tropical parasites (hookworm and malaria) discusses this. The burden of these diseases weighs heavily on children, who are listless and anemic as a result. This can stunt growth and hold back learning (shifting MB down). Historical experience shows that hundred-year-old technology can be used to combat these parasites.
* Supplement nutrition at critical junctures. Poor nutrition can also hold back learning, especially if certain micro-nutrients are missing during critical periods in growth. Advances in this area, such as that of Nevin Scrimshaw, has shown remarkable impacts using very cheap supplements.
* Improve the quality of school. This sounds daunting in principle, but there are small, do-able things that could have large effects. For example, teacher absenteeism is a big problem in the developing world, in part because schools are in far flung places where monitoring is hard. But Esther Duflo, Rema Hanna and Stephen Ryan found experimentally that teachers are more likely to show up (and teach!) if you give them a cheap disposable camera and tie their salary to showing up in a picture of the class each day.

To sum up, there are potentially big gains from improving the quality of the childhood environment, but we should not expect so much from simply jacking up the quantity of time in school.