Improving Equitable Electric Vehicle Adoption Study (IEEVAS)

How to cite

Garg, Teevrat; McCord, Gordon; Bahena, Mexico Alberto Vergara; Muehlegger, Erich; Rapson, David S; Gillingham, Kenneth. 2026. "Improving Equitable Electric Vehicle Adoption Study (IEEVAS)." AEA RCT Registry. AEARCTR-0018940. https://doi.org/10.1257/rct.18940-1.0

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

Teevrat Garg* , Gordon McCord , Mexico Alberto Vergara Bahena , Erich Muehlegger , David S Rapson , Kenneth Gillingham

* Primary investigator

Registry ID
AEARCTR-0018940
DOI
https://doi.org/10.1257/rct.18940-1.0
First registered
Last updated
Status
In development
Trial dates
2026-05-27 – 2028-07-15
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/18940

Abstract

Subsidies for pro-social behaviors—such as adopting electric vehicles (EVs), heat pumps, or other decarbonization technologies—are frequently criticized for being non-additional and regressive. Higher-income households tend to claim subsidies they would have used anyway, while lower-income households often remain constrained by financial, informational, and behavioral barriers. To design more equitable and cost-effective programs, policymakers require evidence not only on how many low- and moderate-income (LMI) households would adopt these technologies at different subsidy levels, but also on how the structure and salience of subsidy offers affect purchase intent, uptake, and the distribution of benefits. We partner with an electric utility and car dealerships to implement a large-scale, two-stage randomized controlled trial across approximately 100,000 LMI households to address these gaps. Combining surveys, negotiated vehicle prices, and vehicle registration records, the study will provide policymakers with experimentally estimated demand curves for EV subsidies among LMI households, evidence on behavioral responses to alternative incentive structures, and the first experimental measures of dealer pass-through in EV subsidy programs.

Experimental design

This study evaluates how program design affects participation in and take-up of electric vehicle incentives among low- and moderate-income households. The trial has three main stages, each involving a distinct randomized comparison between treatment and control groups.

In the first stage, households are randomly assigned to receive different outreach messages describing alternative incentive amounts and allocation rules (treatment groups), while some households will not receive them (control group). This randomization allows us to estimate how program framing affects application behavior, applicant selection, and demand for the program. Applicants who choose to apply complete an intake form that includes demographic, transportation, and charging-access questions, as well as an incentive-compatible preference-elicitation exercise for an alternative rebate.

In the second stage, eligible applicants are randomly assigned either to receive an additional subsidy top-up (treatment group) or to receive only the baseline incentive for which they qualify (control group). Using application records, dealership data, and administrative data, we will estimate the effect of the top-up on vehicle purchase and leasing decisions, including whether households acquire a vehicle and the type of vehicle acquired.

In the third stage, among households receiving incentives, applicants are randomly assigned to disclose subsidy information during the vehicle transaction process, while in the control group information will not disclosed. Comparing negotiated vehicle prices across these groups will allow us to estimate subsidy pass-through and the extent to which incentives are captured by consumers versus dealerships.

Together, these stages provide evidence on participation in subsidy programs, the responsiveness of vehicle purchases to financial incentives, and the incidence of EV subsidies.

Design details

Randomization method
Randomization done in office by a computer.
Randomization unit
The unit of randomization is the household, since they will be invited to participate and we will allow one program application per household.
Clusters
Around 100,000 households will be invited to apply, but only 200 households will receving the rebate and top-ups.
Observations
Around 100,000 households will be invited to apply, but only 200 households will receving the rebate and top-ups.
Arms
In Stage 1 we expect a 33/33/33% split between two main treatment arms and control. In Stage 2, 2/3 of the incentive winners will get a top-up. In Stage 3, 50% of top-up receivers will also get a treatment information.
Primary outcomes
Application Take-Up ( whether a household initiates and completes an application), Willingness-to-Accept (BDM) an alternative rebate, Vehicle puchase/lease, Subsidy Pass-Through

Cited by

No citation data available for this registration yet.

Keywords