Mobility Choices of Employees in a Mobility Budget Scheme – A Randomized Field Experiment

How to cite

Habla, Wolfgang; Wagner, Ulrich; Gessner, Johannes. 2021. "Mobility Choices of Employees in a Mobility Budget Scheme – A Randomized Field Experiment." AEA RCT Registry. AEARCTR-0008470. https://doi.org/10.1257/rct.8470-1.3

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

Wolfgang Habla* , Ulrich Wagner , Johannes Gessner

* Primary investigator

Registry ID
AEARCTR-0008470
DOI
https://doi.org/10.1257/rct.8470-1.3
First registered
Last updated
Status
Completed
Trial dates
2021-04-01 – 2023-01-31
Countries
Germany ()
JEL codes
C93, M14, Q54, R41, R48
Registry record
http://www.socialscienceregistry.org/trials/8470

Abstract

Reducing CO2 emissions from the transport sector is an increasingly important policy objective for governments and companies. To allow policymakers to achieve their objectives at a lower political and economic cost, a better understanding of the effectiveness of so called “soft” interventions such as targeted information provision, social norm activation and moral suasion in the transportation domain is needed.

This study will analyze the effects of social comparisons and moral suasion on transport mode choice in a field experiment. We team up with a multinational corporation which provides its German employees with the option to participate in a novel corporate benefit scheme. In this scheme, participants are provided with an individual mobility budget they can use for private transportation expenditures. This mobility budget allows us to observe expenditures for different transport modes over a two-month intervention period and a two-month post-intervention period for approximately 465 participants. In our field experiment, participants will receive e-mails with treatment messages. Treatment group 1 (the control group) will receive a placebo text with information that is not related to transport mode choice. Treatment group 2 will receive a social comparison message contrasting the participants’ individual transportation behavior with the average in a peer group. Treatment group 3 will receive the same social comparison part as in group 2, plus an additional moral suasion part (in the same message). We will use the data from the mobility budget to analyze the effects of these treatments on the use of different transport modes.

Experimental design

The sample contains all participants in a pilot for mobility budgets at our partner company. The participants are employees at a multinational company working at the companies’ German business locations. They selected into participation in the pilot. In the pilot, participants receive a budget that they can use for private mobility expenses using different modes of transportation. We observe the participants’ expenditures for different transport modes on a daily resolution. Starting in November 2021, we will try to alter participants’ expenditures for public transport (and thus all other categories, as well), by sending emails on a fixed day of the week every other week over a treatment period of two months. Expenses made by household members of the participants will not be considered.
There are 3 treatment arms: treatment 1 serves as a control and receives only a placebo message without information related to transport mode choice. Treatment 2 is intended to activate a social norm with a social comparison message. Treatment 3 combines a moral appeal related to climate change with the social norm message in Treatment 2. We observe transportation expenditures for the 7 months prior to the treatment period and for 3 months after the treatment period.

Design details

Randomization method
Stratified randomization: We will split the sample into strata based on previous transportation expenditures observed in the mobility budget scheme and on the business locations the participants work at. Within the strata, we will randomly assign 1/3 of the participants into each treatment arm. Randomization is done in R.
Randomization unit
Employee participating in the corporate benefit scheme
Clusters
No clustering
Observations
Approximately 465 employees.
Arms
The sample is split with equal probability into the three treatment arms. Each arm will contain roughly 1/3 * 465 participants = roughly 155 participants per treatment arm.
Primary outcomes
Outcomes in the treatment period: - Individual public transportation expenditures – total per time unit [€/time unit] and changes compared to average pre-treatment time unit [Δ€/ time unit], daily, weekly, monthly, over the treatment period - Individual public transportation expenditure shares – share during time unit [€ spent on public transportation during time unit/€ spent in the mobility budget during time unit], daily, weekly, monthly, over the treatment period - Individual public transportation use – count of expenditure items per time unit [Number of expenditure items/time unit], daily, weekly, monthly, over the treatment period - Individual expenditures for car-related mobility – total per time unit [€/time unit], daily, weekly, monthly, total over the treatment - Individual expenditure shares for car-related mobility – share during time unit [€ spent on car-related transportation during time unit / € spent in the mobility budget during time unit], daily, weekly, monthly, over the treatment period - Individual car-related mobility use – count of expenditure items per time unit [Number of expenditure items/time unit], daily, weekly, monthly, over the treatment period - Analogous outcomes for micro-mobility Outcomes after the treatment period: We will analyze all outcomes of the treatment period also for the two-month period following the treatment.
Secondary outcomes
We use the results of a survey which is voluntary among participants and measure the following outcomes for part of the sample: - Intention to use a transport mode most frequently within the mobility budget. (Answer to the survey question: Which of the following means of transport do you intend to use most frequently with the mobility budget in the next six months under the currently given circumstances?) - Positive feelings from acting environmentally-friendly. (Answer to the survey question: It makes me feel good to act environmentally friendly. (Agree?)) - Importance of environmental concerns for transportation decisions. (Answer to the survey question: Environmental concerns play no role for my transport mode choice.) - Importance of behavior observed in the social environment for own decisions (Answer to the survey question: I consider the behavior of my social environment (e.g. Colleagues) when making my own decisions) - Perceived expectations by the social environment with respect to conformity in transportation decisions (Answer to the survey question: My social environment (e.g. Colleagues) expects me to conform with their transport mode choice) - Perceived expectations by the social environment with respect to environmentally-friendly behavior. (Answer to the survey question: My social environment (e.g. Colleagues) expects me to act environmentally friendly)

Papers from this trial

  1. Gessner, Johannes and Habla, Wolfgang and Wagner, Ulrich J., Can Social Comparisons and Moral Appeals Increase Public Transport Ridership and Decrease Car Use? ( 2023). ZEW - Centre for European Economic Research Discussion Paper No. 23-003, Available at SSRN: https://ssrn.com/abstract=4350517 or http://dx.doi.org/10.2139/ssrn.4350517 https://dx.doi.org/10.2139/ssrn.4350517

Cited by

No citation data available for this registration yet.

Keywords