Impact of ESG ratings on consumer behavior

How to cite

Ba, Yuhao; Liu, Tong. 2025. "Impact of ESG ratings on consumer behavior." AEA RCT Registry. AEARCTR-0016755. https://doi.org/10.1257/rct.16755-1.0

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

Yuhao Ba* , Tong Liu

* Primary investigator

Registry ID
AEARCTR-0016755
DOI
https://doi.org/10.1257/rct.16755-1.0
First registered
Last updated
Status
In development
Trial dates
2025-09-22 – 2025-10-22
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/16755

Abstract

Environmental, Social, and Governance (ESG) issues have gained significant prominence in guiding policy development, corporate strategies, and investment decisions. Prominent rating systems -- such as S&P Global’s Dow Jones Sustainability World Index and Bloomberg ESG Ratings -- aim to capture firms’ environmental and social performance, yet they often diverge because of differences in methodology, benchmarks, and theoretical foundations. These inconsistencies raise important questions about how consumers interpret ESG ratings, and whether such ratings influence their choices. Despite growing attention to ESG, empirical evidence on consumer responses remains limited and fragmented.

This paper investigates how companies' ESG ratings influence consumer behavior through an online experiment. We will recruit 1,500 adults to participate in an incentivized survey. Participants will be asked to examine companies with different ESG ratings, and for each company, report their willingness to pay for (i) a gift card that can be used to purchase the company' products and (ii) their willingness to donate to a non-profit initiative sponsored by the company. These two outcome measures will be incentivized. The survey will also include demographic questions, elicit views on environmental issues, and measure behavioral parameters (e.g., risk preference, time preference, and honesty) to explore heterogeneous treatment effects. Finally, we will simulate a major environmental scandal involving a randomly selected company and allow participants to revise their responses, enabling us to examine the extent to which ESG ratings affect a company's resilience to negative news shocks.

Intervention

We will randomly vary the ESG information shown to participants about food & beverage companies.

50% of the participants will be assigned to a Simple condition, which shows only the overall ESG score and grade. These participants will evaluate two companies, shown in a random order: one with an average ESG rating (BBB) and one with a high rating (AA).

The other 50% will be assigned to a Detailed condition, which shows the overall ESG score and grade together with the individual scores for E, S and G. These participants will evaluate six companies in a random order. The six companies differ along two dimensions:
-- Rating: Average (BBB) vs. High (AA)
-- Relative strength: Highest E score vs. Highest S score vs. Highest G score

Experimental design

This is an online experiment in which 1,500 subjects will be recruited to complete a survey. The survey consists of five parts:

1. Background views
-- We ask pariticipants' views on topics such as the environment, climate change, and ESG

2. Behavioral measures
-- We elicit participants' risk preferences, time preferences, honesty, and consumption and donation behaviors

3. Main experiment
-- Participants are randomized into either the Simple or Detailed condition (see Interventions above). They will evaluate food and beverage companies with different ESG ratings. For each company, they will be asked to report (a) their willingness to pay for a $20 gift card that can be used to purchase the company's products, and (b) their willingness to donate to a non-profit initiative sponsored by the company.

4. Scandal
-- We simulate a major environmental scandal involving a randomly selected company, and allow participants to revise their responses.

5. Demographic questions
-- We collect demographic information

Design details

Randomization method
Randomization done by Qualtrics
Randomization unit
Individual
Clusters
N/A
Observations
1,500
Arms
750 in Simple condition 750 in Detailed condition
Primary outcomes
1. Willingness to pay for a $20 gift card that can be used to purchase a company's products 2. Willingness to donate to a non-profit project sponsored by a company (between $0 and $1)

Cited by

No citation data available for this registration yet.

Keywords