Social information and waste disposal

How to cite

Bonan, Jacopo; d'Adda, Giovanna; Cattaneo, Cristina; Galliera, Arianna; Tavoni, Massimo. 2021. "Social information and waste disposal." AEA RCT Registry. AEARCTR-0007834. https://doi.org/10.1257/rct.7834-1.0

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

Jacopo Bonan* , Giovanna d'Adda , Cristina Cattaneo , Arianna Galliera , Massimo Tavoni

* Primary investigator

Registry ID
AEARCTR-0007834
DOI
https://doi.org/10.1257/rct.7834-1.0
First registered
Last updated
Status
On going
Trial dates
2020-07-01 – 2021-10-31
Countries
Italy ()
Registry record
http://www.socialscienceregistry.org/trials/7834

Abstract

We assess the impact of a social information program on the disposal of unsorted waste. We study a program implemented by an Italian multi-utility using an hybrid Pay-as-you-Throw collection system. Households are assigned a maximum number of solid waste units yearly and pay a fixed amount when the quantities are below such limit. For each unit exceeding the cap, customers pay per unit. We randomize customers into a treatment group which receives quarterly reports including information over the volume of unsorted waste compared to the average waste of similar customers living in the city, and a control group which receives nothing. Within the group of treated households receiving the report, half receives the standard report. The other half receives the report along with a clear reference to the disposal cap. This information allows treated customers to keep track of their performance against the cap, which is calculated on a yearly basis. We measure the main and heterogeneous treatment effects on the volume of unsorted waste and on the probability of passing the yearly threshold.

Experimental design

We evaluate the impact of a social information campaign through the design and launch of "Opower-style"home waste reports. The report kicked-off in July 2020 and is delivered by post and email to customersevery quarter, after the waste bill. The reference period of the report is the same of that reported in the bill.The report includes the following elements: 1. static neighbour comparison, 2. Cumulative solid waste disposal, 3. Dynamic feedback, 4. Access to waste collection centers.
The experimental design relies on the random assignment of two thirds of eligible customers to a treatment group which receives the report and one third to a control group, which does not. Within the group of treated households receiving the report, half receives the standard report as described above. The other half receives the report, as described above, along with a clear reference to the disposal cap, displayed in the cumulative solid waste disposal section. This information allows treated customers to keep track oftheir performance against the cap, which is calculated on a yearly basis. These customers also visualize an alert if the volume disposed in the quarter exceeds one fourth of the yearly cap assigned.

Design details

Randomization method
We follow a stratified individual level randomization procedure, to maximize ex-ante balance across the three experimental groups along a battery of important observable characteristics. Strata are obtained from the combination of the following variables: - The disposal caps for unsorted waste in 2020 - The presence of any kind of benefit/deductible in terms of higher cap, due to the presence of children or health needs in the household (at the time of data extraction, i.e. April 2020) - House size above the median - Access to waste collection centres in 2019 - Having a valid e-mail access (hence possibly receiving the report by email vs by post) We exclude strata with less than ten observations. We end up with 96 strata. Within each stratum, we sort customers by baseline unsorted waste disposal volume (in the period April 2019-March 2020) and assign adjacent customers to treatments and control group.
Randomization unit
Individual/user level randomization
Clusters
no clusters
Observations
About 54,150 users
Arms
About 18,050 users in each of the three treatment arms (two treatments and one control group)
Primary outcomes
1. Customer volume of solid waste disposal in the month 2. Probability of exceeding the yearly cap for solid waste disposal

Cited by

No citation data available for this registration yet.

Keywords