This study aims to observe the efficacy of HER in eliciting energy-efficient behaviour from the consumer. We introduce several treatment interventions to a sample of residential electricity customers in Dhaka city. The interventions and both administrative and follow-up data collection will aim to check the heterogenous behavioural implications of the interventions.
We are implementing a multi-arm randomised control trial (RCT in partnership with a government-regulated electricity utility provider Dhaka Electricity Supply Company Limited [DESCO] in Bangladesh. The study will draw a sample of about 2000 households from the residential customer pool of one of DESCO.
Households in the experimental population are randomly assigned into three treatment groups: one placebo-control (C), and two different versions of home energy reporting (HER-1 and HER-2). The sample size of 650 households (rounded) for each of the treatment groups including the control is determined based on ensuring the required statistical power.
We are implementing the following interventions:
>[Placebo-Control: C] We randomly assign 700 households to a placebo-control group. The home in this group is receiving information on what energy-saving behaviours one can adopt. There will be twelve such tips developed through literature search, prior campaign materials found online and finalized after an expert consultation. Additionally, the customers in the placebo group will receive information on their own energy use WITHOUT any comparisons with any other users or customers.
>[Home Energy Report Version 1: HER-1] In this arm of intervention, we will generate monthly reports of energy use using administrative data on electricity meter readings and share with 700 customers. These reports will include (a) the same energy-saving tips as the placebo-control groups; (b) additionally the customers will receive information on their own monthly energy use (in kwh) with further comparisons with (i) other customers within the same neighbourhood and (ii) average use of the customers from the bottom quintile according to the energy use within the same neighbourhood. The neighbourhood is defined by the customers who are connected to the same transformers.
>[Home Energy Report Version 2: HER-2] Behavioral research shows that nominal or monetary comparisons, instead of real units or quantity, can affect customers consumption plans differently. Therefore, in this version of HER, we will make salient how much money the customer is being paid for their electricity consumption both in absolute terms and also relative to the reference groups. The intervention of HER-2, we generate a report similar to HER-1 and share with 700 customers.