Increasing policy acceptance of road pricing policy

We develop an ad-hoc choice experiment to understand what factors are most important for individual acceptance of an innovative transport policy: road pricing. In the choice experiment people are asked to choose between two slightly different road pricing policies and the status quo (no road pricing). We implement the experiments in two cities, Madrid and Oslo.

In the first part of the questionnaire, we implement a battery of questions to record individuals’ characteristics and their transport habits and preferred transport mode.

In the second part, we run the main choice experiment. Participants are presented with a choice several times. Each time participants have to choose between two types of road pricing schemes and the status quo. The two road pricing schemes differ with respect to a set of attributes (prices and use of revenues). The participant choice will be used to estimate consumer (i.e. traveler) preferences as well as preferences for different policies.

In the choice experiment, we test the following attributes:
• Preferences for the status quo (road tolls);
• Price effects with emphasis on actual car use (time and place)
• Preference for Electric Vehicle
• The use of tax revenues

These attributes of the policies will be evaluated for a list of blocks of analysis.
a) Groups of participants: car users vs non-car users.
b) Moreover, these three attributes could be perceived differently if the participants are exposure to specific type of information or not. We identify two different pieces of information:
1. Congestion and pollution
2. Public service
c) Furthermore, there are specific participants’ attitudes that might affect the policy choices. These two are:
1. Perception of inequality: equality drive index
2. Trust in the Government: trust index
d) Localization is also an important factor.
The choice experiment will be further divided in 4 treatments. Participants will be randomly assigned to one of the treatments and will be exposed to different information.

Finally, in the third part, we also explore the role of social norms and attitudes, preferences for redistribution, the role of trust, the socioeconomics traits and, place of residency, that directly influence individuals’ mode of travel and indirectly policy acceptability.

In order to study whether subjects’ mode choice is driven by the presence of social norms, we need to elicit subjects’ Personal Normative Beliefs (PNB), Empirical Expectations (EE) and Normative Expectations (NE). Moreover, in order to study the extent of the influence of social norms, we will also elicit subjects’ mode choice via the strategy method of manipulation of the Empirical and Normative Expectations.

2023-04-16

2024-04-01

Choice of policy vs status quo

effect of prices and EV discount
Effect of the use of revenues

To understand how to facilitate the acceptability of a policy change, we randomly divide responders into three/four treatment groups and a control group. Respondents in each treatment group face a different information, while the control group does not receive any information. Then we implement the same choice experiment for everyone.

• The first treatment is proving information about the negative externalities, such as congestion and pollution, produced by the use of private cars.
• The second treatment presents information about the use of revenues of the tax: the revenues can be devoted to increase the general budget, or destined to specific public necessities.
• The third treatment provide information on public descriptive norms about the acceptance of implementing mitigation policy.
• The fourth treatment, that will be run in Norway, is on road tolls VS road prices (explaining to people what would be the gain of tolls though a numerical example)

Subjects are randomly assigned to one of the four treatments, i.e. Control, Treatment 1, Treatment 2 and Treatment 3. We will check whether the randomization assumption is true by comparing the average and distributions of the socio-economic variables.

Individuals

No

5 arms (1 control and 4 treatments)

We will recruit and start the experiment with 500-1000 subjects.100-200 in each treatment. This number is based on a power analysis calibrated on a pilot analysis run in Madrid in July 2022. The final number of subjects will be influenced by the drop-out rate, which is out of our control and by the recruitment in each country. If needed additional recruitment will be carried out over time.

100-200 individual per arm