Disentangling the costs of meter sharing: experimental evidence from Ethiopia

How to cite

Arega, Tiruwork; Nillesen, Eleonora; Talevi, Marta; Jeuland, Marc; Hailemariam, Robel Seifemichael; Sebsibie, Samuel; Beyene, Abebe; Abate, Gashaw. 2025. "Disentangling the costs of meter sharing: experimental evidence from Ethiopia." AEA RCT Registry. AEARCTR-0015626. https://doi.org/10.1257/rct.15626-4.0

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

Tiruwork Arega* , Eleonora Nillesen , Marta Talevi , Marc Jeuland , Robel Seifemichael Hailemariam , Samuel Sebsibie , Abebe Beyene , Gashaw Abate

* Primary investigator

Registry ID
AEARCTR-0015626
DOI
https://doi.org/10.1257/rct.15626-4.0
First registered
Last updated
Status
On going
Trial dates
2025-06-15 – 2026-12-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/15626

Abstract

A significant share of households in low-income urban areas access electricity from shared meters, which may result in suboptimal electricity use. However, this issue has received little attention from academics and policymakers. While electricity meter sharing may reduce the fixed connection costs, it can contribute to reduced energy efficiency, create agency problems between landlords and tenants (due to imperfect information on different households’ electricity consumption), leading to social conflict (mediated by asymmetries in bargaining power and knowledge), and undermining confidence in service provision by the utility.
More specifically, three potential problems arise from meter sharing. First, households may pay significantly below or above their actual consumption, triggering undesirable responses to price-based energy efficiency policies, or to efforts to promote adoption of energy efficient equipment. Second, the inability to know one's consumption creates trust issues and potential social conflict, as the landlord's interest in bill management may misalign with the tenant's perceptions, needs and actions. Third, suboptimal electricity usage can compromise living standards and slow the energy transition. Thus far, though meter sharing is common in low-income countries and has documented adverse effects on equity under many prevailing tariff regimes, there is a lack of evidence on the challenges that meter sharing poses to bill allocation, energy use behaviour, and related welfare implications in low-income countries. This study aims to provide such evidence, to inform analysis of the social benefits of potential policy interventions that reduce these problems.
The study aims to understand whether introducing a consumption monitoring device resolves information asymmetry in electricity consumption between parties. We aim to answer two questions, using a Randomized Controlled Trial (RCT) with two treatment arms and a comparison group. First, we evaluate the impacts of the consumption monitoring devices on landlord-tenant agency relationship and billing sharing arrangement. Second, we estimate the welfare implications of shared vs. unshared connections, under a volume differentiated tariff structure.

Intervention

Our study explicitly targets multifamily homes, in which the landlord’s household is one of the residents, in low-income neighborhoods in Addis Ababa. Our intervention has two treatment arms and a control group. The first treatment arm provides an electricity monitoring device to each household (i.e., landlords and tenants) living in the same compound. The second treatment arm provides an electricity monitoring device to each household, and in addition, facilitates a landlord-tenant meeting at the time of distribution of those devices at the multifamily home level.

Experimental design

We will randomly distribute multi-family homes having shared meters into three groups. All households in Group 1 will receive an electricity monitoring device and a poster as a usage guideline. All households in Group 2 will receive an electricity monitoring device and the same poster, and they will additionally participate in a purposefully coordinated landlord and tenant discussion. Households in Group 3 will serve as a pure control during the study period; they will not receive monitoring devices or participate in a facilitated discussion.

Design details

Randomization method
We will use a computer to randomize our samples
Randomization unit
Randomization will be done at the multi-family home level.
Clusters
Our samples clusters at multifamily home level. We anticipate enrolling a total of 600 multifamily homes, wherein 300 homes are allocated to treatment groups (150 in each arm) and 300 to control groups, though this is contingent on a listing that will help us determine the number of households living in multifamily homes enrolled in the study.
Observations
2400 households
Arms
On average, we expect each multifamily home to have three tenants and a landlord, yielding four households per home/cluster. The total number of planned observations is 2400 households (600 homes/clusters). We will determine the number of clusters to be used based on a listing, during which the following eligibility criteria will be assessed: multifamily home where the landlord resides in the same home as tenants, all households share a meter, and all households consent to participate in the study.
Primary outcomes
1. Average monthly electricity consumption 2. Average monthly electricity expenditure 3. Trust on service provision 4. Perception of own electricity consumption 5. Bill sharing arrangement 6. Prevalence and nature of landlord/tenant conflicts 7. Appliance ownership and usage frequency 8. Number of new appliances purchased 9. Number of old appliances replaced 10. Number and types of energy efficient practices

Cited by

No citation data available for this registration yet.

Keywords