Spillover, Network, and Equilibrium Effects of Policy Interventions

Programs often affect non-beneficiaries in addition to the individuals directly targeted by the program.  Scaling a program can generate these “spillovers” in at least two ways: first, through network effects, e.g. the benefits of a communications technology depend on how many other people use it, and second, through general equilibrium effects, i.e. changes in market prices like wages that affect all market participants.

The spillovers induced by scale can substantially amplify or undermine the intended effects of a program. One example is the potential for herd immunity that emerges from large-scale public health interventions (Miguel and Kremer, 2004).  In India, rural landowners benefit from weather insurance, but, as more of them take it up, the wages for agricultural workers become increasingly volatile (Mobarak and Rosenzweig, 2014).  Reliance on informal risk sharing anchors workers to their home villages, preventing them from taking advantage of profitable migration opportunities; making formal insurance more available would enable workers to move to better jobs, substantially improving the allocation of labor in the economy (Munshi and Rosenzweig, 2016).  When many households overcome frictions to migrate, even those who stay behind benefit from higher wages in their home communities, thanks to the relative scarcity of labor (Akram, Chowdhury, Mobarak, 2017)

Measuring spillover effects is challenging, but essential to good policy; Y-RISE supports the development and implementation of innovative research designs to overcome those challenges. These designs might include theoretical models disciplined by high-quality microdata, novel data on the structure of pre-existing relationships between beneficiaries and non-beneficiaries, or variation in the scale of a program across markets to detect general equilibrium effects.