Reducing Urban Charcoal Externalities through Solar Technology: Experimental Evidence from Enterprises

How to cite

Alem, Yonas. 2025. "Reducing Urban Charcoal Externalities through Solar Technology: Experimental Evidence from Enterprises." AEA RCT Registry. AEARCTR-0015816. https://doi.org/10.1257/rct.15816-1.0

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

Yonas Alem*

* Primary investigator

Registry ID
AEARCTR-0015816
DOI
https://doi.org/10.1257/rct.15816-1.0
First registered
Last updated
Status
On going
Trial dates
2025-04-14 – 2025-12-31
Countries
Tanzania, United Republic of (Dar es Salaam)
JEL codes
G21, H32, O13, O33, Q23, Q42, Q51
Registry record
http://www.socialscienceregistry.org/trials/15816

Abstract

Charcoal is the key fuel used for cooking by households, enterprises, and other institutions in Africa. Charcoal consumption has serious environmental, climate, and health negative externalities. Previous experimental studies investigating biomass fuel use, including charcoal, exclusively focused on households. Understanding the biomass energy consumption of enterprises and the scope for transition to a cleaner energy source is critical because enterprises consume more than half of the biomass fuel in developing countries. In this project, we investigate the charcoal use behavior of enterprises, more importantly, restaurants and pubs, which consume substantial amounts of charcoal for cooking and boiling water. We collaborate with a reputable solar company and distribute high-cost 200-liter capacity solar water heaters (SWHs) that replace charcoal water boilers on credit with two repayment contracts (payback daily and payback weekly). We identify the impact of SWHs on charcoal consumption, CO2 emission, and enterprise profitability. We also study the credit repayment behavior of enterprises.

Intervention

The intervention randomly provides solar water heaters to a sub-sample of the restaurants on credit under two repayment contracts (payback weekly and monthly).

Experimental design

We randomize 210 enterprises into one of the two treatment groups and a control group. We will provide the control group with full information on the solar water heater (SWH) technology, including the different components, how it works, durability, care, and estimated cost savings from replacing charcoal water boilers with SWHs. We will offer the control group (70 restaurants) the SWHs at the market price (about USD 680) paid upfront. We will provide the first treatment group (70 restaurants) the same information as the control group regarding the SWH technology and will offer the SWHs 20\% the cost of the SWH (USD 680) upfront and the remaining 80\% including the interest in six months payable daily in about 180 small installments. We will offer the second treatment group (70 restaurants) the same information as the control group regarding the SWH technology. We will offer the SWHs 20\% of the cost of the SWH (USD 680) upfront and the remaining 80\%, including the interest payable in six months in 25 weekly payments. Ensol Tanzania Ltd, the solar dealer company we collaborate with, will use the local online cash transfer platform (M-Pesa) to collect the repayment from enterprises. We will collect panel data from the treatment and control groups before and after the interventions to maximize statistical power. The key hypotheses we aim to test are if uptake differs by the type of contract and whether uptake of the SWHs has an impact on the outcome variables of interest.

Design details

Randomization method
We use a Stata randomization generator.
Randomization unit
The randomization units are restaurants.
Clusters
35 clusters.
Observations
For a previous information nudge experiment, we collected baseline data from 1203 restaurants located in 130 clusters (streets) in Dar es Salaam, Tanzania. Streets are the lowest administrative unit in Tanzania. For the current trial, we used the Stata random generator and selected 210 restaurants (70 for treatment 1, 70 for treatment 2, and 70 for the control) spreading in 35 clusters.
Arms
Treatment 1 = 70 enterprises, Treatment 2 = 70 enterprises, and Control = 70 enterprises. We will collect two to three rounds of data before intervention and two to three rounds after intervention, which will give 840 - 1260 observations.
Primary outcomes
The first outcome variable is restaurant uptake of solar water heaters and whether it varies by credit repayment scheme (contract). The second key outcome variable is charcoal consumption and restaurants' corresponding carbon dioxide (CO2) emissions.
Secondary outcomes
The secondary outcome variable of interest is the energy cost to enterprises for water boiling, which is measured by the quantity of each fuel used for water boiling times the unit cost.

Cited by

No citation data available for this registration yet.

Keywords