Chris Blattman and Bernd Beber are looking at the economics of child soldiering.  Why do rebel armies, such as the Lords Resistance Army in Uganda, recruit adolescents?  According to Blattman and Beber, because younger children are ineffective soldiers; and adults are too difficult to indoctrinate. 

Chris Blattman has this graph

What I find interesting about this work is that from premises which sound intuitively plausible, Chris and Bernd then arrive at policy conclusions that are initially counterintuitive:

• anti-insurgency measures (eg increasing military spending by the government) may increase child soldiering (because it increases the size of army needed by a rebel commander)
• measures to reduce child labour may increase child soldiering (because it reduces alternative options for children)
• increased educational and economic opportunities for children will only reduce child soldiering if those opportunities increase faster for children than for adults
• a good strategy to reduce child soldiering would include “abduction training” – teaching children how to resist indoctrination and to escape if captured (rather as Japanese children learn to deal with earthquakes)

Chris calls himself a political scientist, rather than an economist. But I think this is exactly the sort of work that economists, at their best, should be doing.  This work is a fascinating combination of insights into human motivation and incentives, and the use of quantitative techniques to test whether those ideas correspond to the world we observe.