Publication - Working Paper
From the perspective of urban economics, a core constraint facing rural populations is likely their inability to concentrate purchasing power geographically. This is driven by the fact that (i) most primary export products (agricultural goods) require a lot of land, relative to labour for production, and (2) are exports are typically low-margin commodities. We propose to study the effects of relaxing this constraint through a large infusion of purchasing power. Specifically, we will measure the effects of a very large infusion of cash (over USD $11 million) delivered by the NGO GiveDirectly to eligible households in over 650 villages in western Kenya.
We employ a two-level experimental design measuring both direct and spillover effects of the transfers. While overall half of villages were assigned to receive transfers, treatment intensity was varied experimentally at the next highest administrative level (the sub-location) such that half of sub-locations had 2/3 of their villages treated, while the other half had 1/3 treated. Within treatment villages, Households that meet GiveDirectly’s eligibility criteria – residing in a grass-thatched roof house – receive a one-time, unconditional cash transfer, worth about US$1,000, distributed in 3 payments via M-Pesa, a mobile money platform.
This is a unique opportunity to estimate how aggregate demand shocks affect the structure of an economy both experimentally and at a large scale. Results will provide direct evidence on the impacts of cash transfers and on whether a large, one-time cash infusion can move a region to a permanently higher level of economic activity.
Results could also be generalised to shed light on other cash transfer programmes including national safety net programmes for the impoverished, and for understanding how programmes may vary spatially.
Demand for more rigorous evidence on the impacts of cash transfers is significant and growing. Many developing countries and international aid organisations are considering implementing or increasing the use of cash transfers.
In addition to estimating the direct and spillover effects of cash transfers on household welfare, the unique scale and design of this experiment will enable us to answer a number of basic questions about the relationship between purchasing power and spatial equilibrium. In particular, we aim to characterise:
To address these questions, we are conducting household censuses and surveys; enterprise censuses, surveys and midline phone surveys; market price surveys; and surveys of local administrators and community group leaders. All surveys are geocoded, allowing for matching across survey types and an in-depth study of the economic geography of the region.