Making Sense of Sorting

How to cite

Nielsen, Simon Birk; Sørensen, Anders Adser. 2019. "Making Sense of Sorting." AEA RCT Registry. AEARCTR-0004029. https://doi.org/10.1257/rct.4029-1.0

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

Simon Birk Nielsen* , Anders Adser Sørensen

* Primary investigator

Registry ID
AEARCTR-0004029
DOI
https://doi.org/10.1257/rct.4029-1.0
First registered
Last updated
Status
On going
Trial dates
2019-03-04 – 2019-05-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/4029

Abstract

A field experiment seeking to improve the waste sorting among urban households living in apartment buildings in the municipality of Copenhagen. The study will focus on bio waste and use general waste as an indicator of the general level of waste sorting.
The field experiment takes place over a 9-week period within 5 yards, covering around 900 households.
A treatment group will receive a postcard with information regarding the positive effects of waste sorting (information-based instruments) as well as a sticker reminding the subjects to sort their waste (implementation intentions). In the treatment yards, stickers will be placed on all bio and general waste bins, providing feedback to the households of the current sorting levels compared to the average of the neighbourhood using both descriptive and injunctive messages (social norms). These stickers will be updated with new numbers three weeks after the first implementation. A control group will only be observed.
We expect to observe an increase in the amount of bio waste sorted as well as a decrease in the level of general waste, as treated households engage in waste sorting of not only bio waste but other fractions as well, i.e. plastic, metal, paper etc.. The control group is expected to have a relatively constant level of waste sorting.

Experimental design

The field experiment will be conducted in 5 yards across the municipality of Copenhagen for 9 weeks, covering a total of around 900 households in apartment buildings. The yards have been chosen according to when the trans bins are being emptied by the municipality. Having bio waste and general waste emptied the same day ensures precise measurements while making it possible to compare the different years. This requirement has therefore been the determining factor when choosing which yards to include in the field experiment. Alas, the selection of the 5 yards is assumed to be random.
The yards have been assigned to a treatment and control group according to geographical factors making sure that the amount of observations (number of bio and general waste bins as well as number of households expected to use the sorting stations) is the same across the two groups.
The treatment group will receive a postcard with information regarding the positive effects of sorting waste (information-based instrument) as well as a sticker intended to be used to remind themselves to sort their waste (implementation intention). Furthermore, a sticker will be placed on all bio and general waste bins in the yards, informing the households of their current level of sorting compared to the average of the neighbourhood as a whole using both descriptive and injunctive messages (social norms). After three weeks, these stickers will be updated with new averages, giving the households feedback on their progression.
The field experiment will take place over 9-10 weeks where measurements will be take twice a week (general waste is collected twice, bio waste only once).
The first three weeks will serve as a baseline. Then the intervention described above will be implemented followed by three more weeks of measurements. Hereafter, the stickers on the waste bins will be updated, and we will measure for three more weeks.

Design details

Randomization method
Randomization of sample size is done according to the municipality's official plan covering which trash bins are emptied when. Randomization between control and treatment is done based on geographical factors.
Randomization unit
Yards
Clusters
5 yards with a total of 21 sorting stations, each containing several waste bins
Observations
21 sorting stations covering 32 bio waste bins and 78 general waste bins
Arms
3 yards control (12 sorting stations), 2 yards treatment (9 sorting stations)
Primary outcomes
Volume of bio waste and general waste

Cited by

No citation data available for this registration yet.

Keywords