The Effect of Peer Comparisons on the Behavior of Facilities Regulated Under the Clean Water Act: A replication

How to cite

Ferraro, Paul; Earnhart, Dietrich. 2025. "The Effect of Peer Comparisons on the Behavior of Facilities Regulated Under the Clean Water Act: A replication." AEA RCT Registry. AEARCTR-0017140. https://doi.org/10.1257/rct.17140-1.0

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

Paul Ferraro* , Dietrich Earnhart

* Primary investigator

Registry ID
AEARCTR-0017140
DOI
https://doi.org/10.1257/rct.17140-1.0
First registered
Last updated
Status
On going
Trial dates
2024-03-04 – 2026-12-31
Countries
United States of America ()
JEL codes
Q5, D9
Registry record
http://www.socialscienceregistry.org/trials/17140

Abstract

Using estimation procedures from a pre-analysis plan and data from a randomized controlled trial with 328 municipal wastewater treatment facilities in Kansas (https://osf.io/6u7t8/), Earnhart and Ferraro (2021) estimated the average effect of a peer (social) comparison letter on a facility’s reported biological oxygen demand effluent. The peer comparison letter contrasted how the facility’s discharge-to-limit ratio under the U.S. Clean Water Act compared to the ratios of other facilities in the state. In the 18-month post-intervention period, the letter recipients reported discharge-to-limit ratios that were 8% lower, on average, than non-recipients (95% CI [-15%, -1%]). For the subgroup of monthly reporters, the letter recipients reported discharge-to-limit ratios that were 11% lower, on average, than non-recipients (95% CI [-22%, -1%]). In a new trial, The Nature Conservancy aims to replicate the Earnhart-Ferraro trial using a larger, more heterogeneous sample of monthly reporting facilities from multiple states permitted under the National Pollutant Discharge Elimination System.

Intervention

The intervention is a peer (social) comparison letter that contrasts how a polluting facility’s discharge-to-limit ratio for biological oxygen demand effluent levels under the U.S. Clean Water Act in the prior fiscal year compares to the ratios of other facilities in its state.

Experimental design

National Pollutant Discharge Elimination System (NPDES) facility contact names were blocked-randomized to either a treatment (intervention) and or a control (no intervention) group in approximately equal proportions.

Design details

Randomization method
Block randomization on computer using Stata command - block_ra - . The blocking covariates, measured at the facility level, were the state, the sectoral subgroup, and the quartile of average FY23 BOD discharge-to-limit ratio. For NPDES contacts associated with more than one facility: the average BOD values of the facilities associated with the contact was used to obtain their quartile for blocking, the modal sectoral grouping of the facilities associated with the contact was used to obtain their sectoral group for blocking, and the modal state of the facilities associated with the contact was used to obtain their state for blocking.
Randomization unit
NPDES facility contact name
Clusters
4,602 facility contact names
Observations
5,076 facilities (most contact names are associated with only one facility)
Arms
2,503 facilities in the treatment group, 2,573 facilities in the control group
Primary outcomes
The outcome variable (dependent variable) is the natural logarithm of the ith facility’s biological oxygen demand (BOD) discharge-to-limit ratio in month t. The discharge-to-limit ratio is the ratio of the facility’s reported BOD discharge in its Discharge Monitoring Report to the facility’s permitted BOD limit. See the trial's Pre-Analysis Plan for more details.

Cited by

No citation data available for this registration yet.

Keywords