REDGAS: A RANDOMIZED EVALUATION OF THE DEMAND FOR LPG STOVES AND OF THE ASSOCIATED EFFECTS ON HOUSEHOLD AIR POLLUTION, GREENHOUSE GAS EMISSIONS AND WELL-BEING IN BURKINA FASO

How to cite

Vreyer, Philippe De; Djemaï, Elodie; Thivillon, Thomas. 2019. "REDGAS: A RANDOMIZED EVALUATION OF THE DEMAND FOR LPG STOVES AND OF THE ASSOCIATED EFFECTS ON HOUSEHOLD AIR POLLUTION, GREENHOUSE GAS EMISSIONS AND WELL-BEING IN BURKINA FASO." AEA RCT Registry. AEARCTR-0005024. https://doi.org/10.1257/rct.5024-4.0

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

Philippe De Vreyer* , Elodie Djemaï , Thomas Thivillon

* Primary investigator

Registry ID
AEARCTR-0005024
DOI
https://doi.org/10.1257/rct.5024-4.0
First registered
Last updated
Status
On going
Trial dates
2019-11-21 – 2023-03-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/5024

Abstract

Recent epidemiologic studies have shown that household air pollution from cooking on inefficient biomass stoves is a major public health concern which causes between 2.8 and 4.2 million premature deaths every year (WHO 2014). LPG is one of the cleanest burning cooking fuels and has been shown to have a low climate impact throughout its life cycle. This research project aims to quantify the health and socioeconomic impacts of access to LPG for low income households in developing countries. It will be implemented in partnership with NGO Entrepreneurs du Monde (EdM) in a country where it is currently conducting a LPG promotion program.

A randomized controlled experiment will be conducted in 2022. It will involve a random sample of 823 households who will be assigned to one of three groups :
1 A ‘subsidy group’ in which households will be offered a 38% discount on the purchase of a LPG kit including a stove and a 6kg cylinder;
2. A ‘credit group’ : the households assigned to this group will receive a sales offer enabling them to purchase the same kit in three installments (but at the standard market price);
3. A ‘comparison group’ where households will not receive any offer.

29% of the households will be assigned to the 'subsidy group', 29% to the 'credit group' and 42% to the control group.

The research team will collect a combination of standard self-declared survey data and field measurements both before the intervention and after the intervention (6 to 10 months after the intervention). Electronic Stove Usage Monitoring Systems (SUMS) will be installed on LPG stoves to record their usage frequency during the whole study. Individual exposure to PM2.5 will be measured with ambiant air pumps and ‘light scattering sensors’ (24h gravimetric measurement) at baseline, midline and endline. The Kitchen Performance Test (KPT) Protocol will be used to assess the fuel consumption of the households over 24 hours to 3 days.

The collected data will allow to answer a series of questions concerning the determinants of LPG adoption (household characteristics, budget constraint, beliefs about cooking with wood or charcoal) and its impact on exposure to indoor air pollution, health and time use (women and children).

Intervention

The project aims at identifying, first, what determines the household choice of fuel for cooking and the kind of incentive that may be effective to induce the household to change for a clean cooking solution (towards LPG stove), and second, the impact of adopting LPG as a cooking solution on exposure to particulate matter (PM2.5), as well as on women's and children's time use and health.

Experimental design

A randomized controlled experiment will be conducted in 2022. It will involve a random sample of 823 households who will be assigned to one of three groups :
1 A ‘subsidy group’ in which households will be offered a 38% discount on the purchase of a LPG kit including a stove and a 6kg cylinder;
2. A ‘credit group’ : the households assigned to this group will receive a sales offer enabling them to purchase the same kit in three installments (but at the standard market price);
3. A ‘comparison group’ where households will not receive any offer.

29% of the households will be assigned to the 'subsidy group', 29% to the 'credit group' and 42% to the control group.

The research team will collect a combination of standard self-declared survey data and field measurements both before the intervention and after the intervention (6 to 10 months after the intervention). Electronic Stove Usage Monitoring Systems (SUMS) will be installed on LPG stoves to record their usage frequency during the whole study. Individual exposure to PM2.5 will be measured with ambiant air pumps and ‘light scattering sensors’ (24h gravimetric measurement) at baseline, midline and endline. The Kitchen Performance Test (KPT) Protocol will be used to assess the fuel consumption of the households over 24 hours to 3 days.

Design details

Randomization method
Sampled households are randomly selected in three towns. First, the coordinates of "GPS search starting points" are randomly selected by the scientific investigators using a GIS. Once on the field the surveyors have to find the search points and then to follow a procedure to select a household to interview, then to check his eligibility and to obtain his consent to take part to the survey. In each town we use the map of census districts to identify the limits of the city. From the most recent satellite images we then delimit the urban area of high population density as peripheral census districts often include large areas with no of very few dwellings. The resulting areas is then split between 56 "cells" of identical surface. Within each cell a fixed number of households is drawn from the population of households that are located at least 100 meters from the cell's frontier. This way "blocks" of households are spaced from each other by at least 200 meters. The assignment into the three treatment groups is done in office by a computer. The balance checks are done on the main variables observed before the intervention starts.
Randomization unit
The randomization will be at the household level.
Clusters
NA
Observations
821
Arms
239 households receive an offer to buy the LPG stove at a discounted (38%) price 236 households receive an offer to buy the LPG stove at credit (three installments) 346 households do not receive an offer to buy the LPG stove
Primary outcomes
- LPG adoption and usage (ownership of LPG stove, declared frequency of use, monitored frequency of usage, number of cooking events using the LPG stove per week, per month) - Household consumption of various fuels (wood, charcoal, kerozene, gas etc.) - Air pollution
Secondary outcomes
- Time use (women and children) - time spent in domestic chores, school attendance, labor market participation - Health (Self reported health, health expenditures, declared symptoms) - COVID19 infection (TBC)

Cited by

No citation data available for this registration yet.

Keywords