A Natural Field Experiment on (Community) Incentives in Bicycle-share Systems

How to cite

Gosnell, Greer. 2017. "A Natural Field Experiment on (Community) Incentives in Bicycle-share Systems." AEA RCT Registry. AEARCTR-0002216. https://doi.org/10.1257/rct.2216-4.0

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

Greer Gosnell*

* Primary investigator

Registry ID
AEARCTR-0002216
DOI
https://doi.org/10.1257/rct.2216-4.0
First registered
Last updated
Status
Withdrawn
Trial dates
2017-05-22 – 2018-08-31
Countries
United States of America (Atlanta, GA)
JEL codes
D12, L9, Q5, R4
Registry record
http://www.socialscienceregistry.org/trials/2216

Abstract

We have partnered with a major US city that has recently developed its bicycle-sharing system to increase access to healthy and environmentally sustainable transport for its citizens. Historically, the city's transit agencies have used sign-up discounts to incentivize residents to utilize publicly provided transportation options. However, insights from behavioral science and economics have demonstrated that financial incentives may not be the most (cost-) effective means of encouraging individuals to make presumably welfare-maximizing decisions on the margin. Here, we aim to understand the importance of personal and community incentives in motivating individuals to participate in the city’s newly expanded bicycle sharing system, a service that may arguably be underutilized due to both myopia in health-related decision-making and the lack of prices to internalize social damages (e.g., pollution and greenhouse gas emissions) associated with conventional transport options. To discern whether there is a role for altruism and incentives (and assess their interaction), we will randomize the content of mailers that will be sent to approximately 30,000 homes across seven neighborhoods to introduce an expansion of the cycle scheme into those homes’ neighborhoods. Our analysis will focus primarily on differences in initial uptake (i.e. extensive margin) and the extent of participation (i.e. intensive margin). The results will have implications for cities looking to promote engagement with sustainable public transport.

Experimental design

The experiment comprises 29,878 households within a major US City in 2017. In alignment with the SUTVA assumption of non-interference, we assume that individuals who received postcards did not discuss the postcards they received with their neighbors. However, we will investigate within-household spillovers by identifying instances in which two or more individuals sign up within the same household (in cases where we observe and track more than one sign-up at a given single-unit address).

We use address data from public tax records to perform the randomization. As such, apartment unit-level randomization is only possible for units that are owner-occupied. For all apartments where apartment units are not identifiable in tax record data, we randomize at the apartment level so that all individuals within a given apartment block receive the same mailer. To ensure that we achieve balance on location and house type (e.g., single-family, duplex, multi-unit of various sizes) in the allocation of households to the various study groups, we conduct a stratified randomization, blocking first on the neighborhood in which the household resides and second on a categorical variable that captures the number of livable units at the given address (for multi-unit addresses that are not owner-occupied). A vast majority of addresses (87.9%) are identified as single-unit, and another 7.5% are identified as duplexes; these constitute the first two levels of the livable units categorical variable. The remaining categories comprise 3-4 units (1.9%), 5-8 units (1.0%), 9-20 units (1.1%), 21-50 units (0.2%), 51-100 units (0.1%), 101-200 units (0.1%), 201-300 units (0.1%), 300-500 units (0.1%).

Four apartment blocks that were over 500 units were excluded from the experiment for cost purposes, since large clusters provide less benefit in terms of power. We exclude three neighborhoods that had already been exposed to the bike share program during the demonstration phase of the bicycle-share system. We additionally exclude two university neighborhoods where the ability to randomize at the household level is limited and risk of spillovers is likely high. A key component of the analysis is the ability to match billing addresses with mailer recipient addresses, especially for the control group, which does not receive a promo code to allow for tracking. Since many students use their permanent hometown addresses as their billing addresses, there is strong potential for inability to match data outcomes to interventions received.

Design details

Randomization method
Stratified randomization done in Stata by the PI
Randomization unit
We randomize at the address level, where a majority (~88%) of addresses are single-unit (i.e. household-level randomization) and the remaining addresses are used in the stratification to ensure balance across the size of the complex.
Clusters
14,452 clusters (12,701 single-unit addresses; 1751 multi-unit addresses)
Observations
29,878 households
Arms
Control: 4510 households (2404 clusters) Treatment 1 (Prosocial, Own Community Hub): 4735 households (2407 clusters) Treatment 2 (Prosocial, Low-Income Community Hub): 5183 households (2412 clusters) Treatment 3 (Incentive, $5 Discount): 5649 households (2409 clusters) Treatment 4 (Incentive, $8 Discount): 4642 households (2407 clusters) Treatment 5 (Prosocial + Incentive, Low-Income Community Hub + $5 Discount): 5159 households (2413 clusters)
Primary outcomes
Cycle scheme registration and first ride (extensive margin) Intensity of bike share use (i.e. number of rides; intensive margin)

Cited by

No citation data available for this registration yet.

Keywords