|
Field
Last Published
|
Before
October 02, 2017 09:43 AM
|
After
November 30, 2017 09:55 AM
|
|
Field
Intervention (Public)
|
Before
|
After
This pilot, running from October 2017 - February 2018, will consist of randomized household subsidies of energy efficient cookstoves for 200 households residing in Kibera. A randomized, continuous range of subsidies will induce two groups: adopters and non-adopters. Compliance with subsidies is not expected to be perfect, hence subsidies will be used as an instrument for cookstove ownership. Follow-up via SMS with all households will measure energy expenditures post-intervention.
A more complex experimental treatment design will be used for the full-size study later in 2018.
|
|
Field
Primary Outcomes (End Points)
|
Before
- Energy expenditures
- WTP for energy efficiency
|
After
The pilot will primarily focus on the following two outcomes:
- WTP for energy efficient cookstoves
- Energy expenditures post-intervention
The more complex treatment design in 2018 will contain additional outcome variables.
|
|
Field
Primary Outcomes (Explanation)
|
Before
|
After
WTP: elicited through a BDM mechanism.
Energy expenditures: total household spending on charcoal elicited through follow-up SMS.
|
|
Field
Experimental Design (Public)
|
Before
We first prompt potential purchasers to exercise greater attention by calculating expected savings from the stove. We then study the relationship between this inattention problem and traditional uncertainty by cross-randomizing whether participants have access to a trial stove for a week before making their purchasing decision. Then, to understand whether households view energy savings through the lens of mental accounting we randomly allocate cash transfers equivalent to the expected savings and test for differences in consumption responses. Finally, we use high-frequency monitors to estimate how cash transfers induce consumption changes in order to benchmark the household’s welfare gains derived from improved energy efficiency.
|
After
During the pilot, we employ a BDM mechanism to elicit household willingness-to-pay for an energy efficient cookstove, and conditional on household willingness-to-pay, randomize cookstove ownership. Our goal is to enrol 200 households (from a sample pool of 300 recruited households) consisting of 100 adopters and 100 non-adopters. Through an SMS survey we then measure recurring charcoal expenditures to experimentally test for any reduction in energy spending for households in the treatment group. In addition to qualitative information the pilot design will generate two quantitative outcomes: 1) A precise demand curve of cookstoves across a dense distribution of price points, and 2) The treatment effect of cookstove ownership on energy expenditures.
We first prompt potential purchasers to exercise greater attention by calculating expected savings from the stove. We then study the relationship between this inattention problem and traditional uncertainty by cross-randomizing whether participants have access to a trial stove for a week before making their purchasing decision. Then, to understand whether households view energy savings through the lens of mental accounting we randomly allocate cash transfers equivalent to the expected savings and test for differences in consumption responses. Finally, we use high-frequency monitors to estimate how cash transfers induce consumption changes in order to benchmark the household’s welfare gains derived from improved energy efficiency.
|
|
Field
Randomization Method
|
Before
Electronic randomization.
|
After
Randomization done in office by a computer.
|
|
Field
Planned Number of Observations
|
Before
500 households.
|
After
For the pilot: 200 households.
A larger sample will be used for the full study.
|
|
Field
Sample size (or number of clusters) by treatment arms
|
Before
Treatment arms and sample sizes TBD.
|
After
100 cookstove adopters and 100 non-adopters.
|