Rebound in household cooking technologies
How to cite
Vance, Colin; Frondel, Manuel; Bensch, Gunther. 2018. "Rebound in household cooking technologies." AEA RCT Registry. AEARCTR-0002803. https://doi.org/10.1257/rct.2803-1.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
We partnered with the Rwandan governmental agency Energy Water Sanitation Authority (EWSA) in a study that pilots different dissemination strategies for the national cookstove program in Rwanda using a cluster randomized trial. The national cookstove program promotes improved cookstoves (ICS) that consume less firewood than traditional cookstoves. Drawing on a sample of households from treatment and control regions, we estimate the magnitude of the rebound effect, that is, the degree to which energy savings from an ICS are offset by an increase in demand for cooking services.
Intervention
We randomized different subsidies for cookstoves across a sample of 84 communities. These communities were randomly allocated in October 2014 to a control group, a zero-price group, a full price and a half price group (21 communities each).
In the three treatment groups, stove production units and installers were trained in a business-as-usual way. Furthermore, in cooperation with the installers we arranged a marketing event in each of the treatment communities, informing households about the stove’s main advantages and usage instructions. These events took place between May and August 2015. Rural dwellers in the visited communities could make orders for the ICS right after this marketing event, but also on a rolling basis. The three groups only differ in the price at which the stove was offered. The ordered ICS was installed in the weeks following the event. Households in the full and half-prize group could make payments in installments.
Experimental design
We randomized different subsidies for cookstoves across a sample of 84 communities. These communities were randomly allocated in October 2014 to a control group, a zero-price group, a full price and a half price group (21 communities each).
Design details
Cited by
No citation data available for this registration yet.