Electric cooking and sustainable development: experimental evidence from eastern D.R. Congo

Worldwide, around 2.6 billion people remain dependent on biomass fuels for cooking, posing serious environmental costs. Among them, close to 2 billion people now have access to electricity. This opens opportunities, the more so since technological innovations in Electric Pressure Cookers (EPC) have made cooking with electricity a cheaper and faster alternative. Opportunities are especially promising in Eastern Africa, where electricity grids are becoming more reliable and reaching more people, and because EPCs are well-suited for the East African kitchen. We offer experimental evidence assessing the optimal delivery mechanism to maximize the adoption and use of EPCs.

Due to the novelty of the product, the associated information gap and the high upfront cost (around 70usd per EPC), selling the EPC at market price may lead to a low take up. We will test a distribution model with a 100% subsidy provided by a local electricity distributor, Virunga Energies (VE). As the energy distributor would increase its electricity sells, such a high subsidy is cost-effective from
its perspective if EPC use is sufficiently high. Furthermore, we provide a free electricity voucher to a subgroup of beneficiaries to overcome the learning cost of EPC and make the transition easier at the beginning from the perspective of the household. Finally, we cross-randomize this treatment with a focus on social benefits of EPC adoption.

2022-09-05

2023-03-31

1. Usage of the cooker: Monthly electricity consumption proxied by monthly electricity purchases (in usd) from m-6 to m+12

2.1 Charcoal consumption: Average daily weight of ashes

2.2 Charcoal consumption: Monthly spending on charcoal (in usd)

2. 3. Charcoal consumption: charcoal from the park (dummy)

3.1 Consumption: Spending (in usd) on non-energy related items (recall question, simplified consumption module)

4. Pro-environmental motivations: Contribution to a dictator game during which participants will be given the opportunity to contribute up to 5usd to local NGOs.

Average daily weight of ashes: During the first post-intervention survey, a 10L bucket will be distributed to participants. We will ask participants to dispose all their ashes in the bucket. About seven days later, the enumerator will come back to weigh the ashes. The total weight will be divided by the number of days during which the households disposed her ashes in the bucket.

The core intervention (denoted T below) consists of 1,000 randomly selected clients receiving a free Electric Pressure Cooker (EPC). The EPCs are worth around 70usd in Kenya and Tanzania and are not available in Eastern DRC at the start of the experiment. They will be distributed to beneficiaries during demonstration sessions organized by Virunga Foundation staff and extension workers (“ambassadors”). The demonstration sessions will include about 15 to 20 participants per session, and last around 2 hours. Staff, ambassadors, and participants will cook local recipes together. EPCs will be connected to electricity meters to show clients the actual energy usage of the cookers. Participants will share the meals they cooked together, allowing them to taste the food cooked with the EPC. Each participant is assigned to a specific ambassador who will assist and monitor them in the first two months. A second group of 500 randomly selected clients will not receive a cooker, and serve as the control group (C).

We will further randomize EPC beneficiaries in four treatment arms to explore mechanisms of adoption:

2.1.1 EPC with electricity voucher (“V1”): Households in V1 receive a 5usd (20kwh) voucher of free electricity for their meter. This
treatment arm intends to allow beneficiaries to try the cooker for about 25 meals without enduring any monetary cost, hence allowing them to continue buying charcoal in case they are not satisfied with the EPC.

2.1.2 No voucher (“V0”): Participants in V0 receive the EPC during a demonstration session (our core intervention), but do not receive a 5usd voucher of free electricity

2.3.1 Pro-social preferences – Nudge (“N1”): During the demonstration sessions, a module of around 30 minutes will detail to participants the consequences of charcoal production on the conservation of VNP (deforestation and destruction of habitats for flagship species, including Mountain Gorillas and Chimpanzees) and the impact of illegal charcoal production on the revenues of armed militias. This module will be designed with Virunga’s Environmental Education department and conducted by experienced rangers.

To reinforce the impact of the nudge, a small sticker designed by a local artist will be put on theEPC. The sticker will convey a simple message around the benefits of using eCooking instead of illegal
charcoal to protect the environment and promote peace.

2.3.2 Pro-social preferences -Control (“N0”): Participants in N0 receive the EPC during a demonstration session (our core intervention), but not the nudge.

To avoid contamination between treatment arms, all participants in a given demonstration session will be assigned to the same treatment arm ({V0, N0}, {V0, N1}, {V1, N0}, {V1, N1\}), and randomization will be done based on spatial clusters. In addition, ambassadors will also be assigned to a single treatment arm.

Randomization done in office by a computer (R software, cluster randomization). Randomization is stratified by previous electricity consumption (average six months before the intervention, in kwh), and previous charcoal consumption (month before the intervention, in usd)

Clusters of households living within a given distance from each others (around 150 meters) are assigned to the five different treatment arms.

Yes

400 (200 clusters per wave)

1,500 households

1000 beneficiaries:
- 250 receiving only a cooker
- 250 receiving a cooker with 5usd electricity voucher
- 250 receiving an environmental nudge
- 250 receiving a cooker with 5usd electricity voucher and an environmental nudge

500 control households

Based on data from the pilot with 50 cookers, we will be sufficiently powered at 0.8 to measure an estimated treatment effect size below 4kwh of monthly electricity consumption (+0.28sd) with 250 beneficiaries per treatment arm. For charcoal consumption, our study should be powered to detect a 2kgs change in monthly charcoal consumption (5% change) with about 250 participants per treatment arm.

PAP (October 3rd 2022)