Investigating the Effects of Eco-Friendly Service Options on Rebound Behavior in Ride-Hailing

TL;DR

This paper investigates whether eco-friendly service options (EFSOs) on ride-hailing platforms induce rebound-like behavior, using an online within-subject experiment (N=75) that compares five variants (No EFSO plus Minimal, CO$_{2}$ Equivalency, Social, and Gamified). A logistic mixed-effects model reveals that some EFSO designs modestly increase ride uptake, especially as distance grows, while richer eco-feedback variants do not consistently amplify rebound relative to a minimal leaf-icon view. Qualitative analyses highlight that decision-making remains dominated by convenience and comfort, with EFSOs serving as moral-psychological justifications that can legitimize less environmentally favorable choices. The study offers practical guidance for designing EFSOs to curb rebound effects, including transparent communication of benefits and attention to moral-psychological framing, contributing to Sustainable HCI research and the responsible deployment of eco-feedback in online platforms.

Abstract

Eco-friendly service options (EFSOs) aim to reduce personal carbon emissions, yet their eco-friendly framing may permit increased consumption, weakening their intended impact. Such rebound effects remain underexamined in HCI, including how common eco-feedback approaches shape them. We investigate this in an online within-subjects experiment (N=75) in a ride-hailing context. Participants completed 10 trials for five conditions (No EFSO, EFSO - Minimal, EFSO - CO2 Equivalency, EFSO - Gamified, EFSO - Social), yielding 50 choices between walking and ride-hailing for trips ranging from 0.5mi - 2.0mi (0.80km - 3.22km). We measured how different EFSO variants affected ride-hailing uptake relative to a No EFSO baseline. EFSOs lacking explicit eco-feedback metrics increased ride-hailing uptake, and qualitative responses indicate that EFSOs can make convenience-driven choices more permissible. We conclude with implications for designing EFSOs that begin to take rebound effects into account.

Figures (7)

• Figure 1: Visualizations of the ride-hailing variants used in our study (derived from mohanty2023). a) represents a regular gas-powered vehicle. In the No EFSO variant, only gas-powered ride-hailing options exist. b)-d) represent eco-friendly ride options. In b), the eco-friendly ride is shown using a green leaf icon and not additional information (EFSO - Minimal). In c) the raw CO$_{2}$ output is communicated together with more relatable carbon equivalencies (EFSO - CO$_{2}$ Equivalency). In d) the positive environmental impact of the eco-friendly ride is represented using a gamified metric based on "EcoPoints" (EFSO - Gamified). In e) the environmental impact of ride options is communicated using the collective CO$_{2}$ emissions by users within the last week (EFSO - Social).
• Figure 2: The figure shows the interface of our study application. Panel (a) presents the route information section, which displays walking and driving maps along with the meeting location, starting point, distance, and estimated travel duration. Panel (b) shows the choice interface, featuring two ride-hailing options and a button to select walking. Panel (c) displays the reflection interface, where participants were shown the shortest distance at which they switched from walking to ride-hailing for each EFSO variant and were prompted to provide a free-text rationale for that decision.
• Figure 3: A visual process diagram showing a schematic overview of the study procedure as described in \ref{['sec:procedure']}.
• Figure 4: Figure showing four plots with observed probability of choosing ride-hailing over walking at each trip distance. Each plot shows the contrast between No EFSO and either EFSO - Minimal (top-left), EFSO - CO$_{2}$ Equivalency (top-right), EFSO - Gamified (bottom-left), or EFSO - Social (bottom-right). Dots represent the mean probabilities, and error bars indicate 95% confidence intervals. Distances are shown in miles and kilometers.
• Figure 5: Figure showing the log-odds contrasts compared to No EFSO. The dashed horizontal line indicates the no-difference threshold relative to No EFSO, and error bars represent 95% confidence intervals. The panel on the right displays the corresponding odds ratios (OR) along with their 95% confidence intervals, with L denoting the lower bound and U the upper bound.