Overcoming Supply Side Frictions for Distribution

How to cite

Green, Brett; Fuchs, William; Levine, David. 2018. "Overcoming Supply Side Frictions for Distribution." AEA RCT Registry. AEARCTR-0003028. https://doi.org/10.1257/rct.3028-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Brett Green* , William Fuchs , David Levine

* Primary investigator

Registry ID
AEARCTR-0003028
DOI
https://doi.org/10.1257/rct.3028-1.0
First registered
Last updated
Status
Completed
Trial dates
2013-01-01 – 2016-04-30
Countries
Uganda (Africa)
JEL codes
I15, O1, O33, D02, D86
Registry record
http://www.socialscienceregistry.org/trials/3028

Abstract

A large literature examines demand-side barriers to product adoption. In this experiment, we investigate supply-side barriers. We recruited entrepreneurs from BRAC's network of community health providers (CHPs) to distribute solar lights in rural Uganda. We provide entrepreneurs with training and information on solar lights. We then randomized over which arrangement to offer to the entrepreneurs. There were three contractual feature that we varied, one aimed at addressing credit constraints, one aimed at addressing vendor uncertainty, and one aimed at addressing customer uncertainty. We explored how each of these features influence overall sales performance of entrepreneurs.

Intervention

We randomly offered trade credit and the right to return unsold inventory to vendors of solar lights. We also randomly offered a "loaner light" that vendors could lend to customers for free trials.

Experimental design

We randomly offered trade credit and the right to return unsold inventory to vendors of solar lights. We also randomly offered a "loaner light" that vendors could lend to customers for free trials.

Design details

Randomization method
We divided each BRAC branch into 4 zones of several nearby vendors. We randomized zones using the pseduo-random number generator in Excel. Within each zone, we randomly chose one vendor to receive the "loaner" solar light.
Randomization unit
We randomized zones for credit and right to return. We randomized vendors for who would receive the loaner light first.
Clusters
We planned 8 branches with 4 zones per branch = 32 zones (the unit of randomization).
Observations
We anticipated 4 vendors per zone * 32 zones = 128 vendors.
Arms
16 zones with credit and right to return (that is, consignment) 16 zones with credit only 16 zones with right to return only 16 zones with neither
Primary outcomes
Orders of solar lights by vendors

Cited by

No citation data available for this registration yet.

Keywords