Expanding Governance as Development : Evidence on Child Nutrition in the Philippines

Worldwide, extreme poverty is often concentrated in spaces where people and property are not safe enough to sustain effective markets, and where development assistance is dangerous - and might even induce violence. Expanding governance by coercively taking control of territory may enable markets and development programs, but costs to local residents may exceed benefits, especially if that expansion is violent. We estimate for the first time whether a large counterinsurgency program improves welfare. We exploit the staggered roll-out of the Philippine "Peace and Development Teams" counterinsurgency program, which treated 12% of the population between 2002 and 2010. Though treatment temporarily increased violence, the program progressively reduced child malnutrition: by 10% in the first year, and by 30% from year three onwards. Improved nutritional status was not due to increased health and welfare expenditures, but instead to improved governance. Treatment effects are comparable to those of conventional child health interventions, though conventional programs are likely infeasible in this setting. Rebels apparently react to treatment by shifting to neighboring municipalities, as malnutrition worsens there - with statistically significant 'treatment' effects of similar size. Thus overall program effects are close to zero. These findings invite an evidence-based discussion of governance expansion, an extensive margin of development