New England mileage fee survey
Data files
Aug 30, 2023 version files 735.97 KB
-
NE_MF_Survey_Metadata_2023.csv
-
README.md
Abstract
Fleet diversification and increases in energy efficiency continue to weaken the revenue-generating ability of motor fuels taxes (colloquially, “gas taxes”), which are a large source of funding for transportation projects. While alternative funding schemes are necessary, consensus amongst policymakers is lacking and public acceptance of changes to the gas tax is low. We surveyed residents of Vermont, Maine, and New Hampshire to gauge understanding of and support for a mileage fee and a flat fee as potential replacements for the gas tax. Throughout the survey, respondents were provided information and learning opportunities to “myth bust” common misconceptions about the gas tax and the potential policy alternatives. We find that, before education, respondents knew very little about how the current gas tax works and showed minimal support for the proposed policy alternatives. Post-education, support for mileage fees increased by 11%, and the impact of the education was statistically significant in increasing policy support. Additional regression models revealed that while perceptions of fairness may not be easily changed with education in a survey format, presenting respondents with personalized cost estimates was a highly effective way to increase policy support. Overall, we find responding to common public concerns with up-to-date and non-biased information within a relatively simple learning experience can cause substantial changes in policy support. Our findings offer an avenue to understand how support for gas tax alternatives varies amongst different groups of people and the role that education can play in increasing policy support in the face of widespread misconceptions.
Methods
We created an internet-based survey to gather public opinion on a $0.015 per mile travelled fee (mileage fee) and a $220 per year per vehicle fee (flat fee) to replace state gas taxes in Vermont, New Hampshire, and Maine. The survey was fielded between May 6th and June 3rd of 2022 using Qualtrics paid survey panelists. Each state was surveyed to ensure 210 usable responses per state. A total of 658 complete responses were collected.
In the survey, respondents were presented with voting opportuntities (Do you support replacing the gas tax with a mileage fee? Do you support replacing the gas tax with a flat fee?), followed by educational treatments. The order was as follows: Voting Opportunity 1, Education Treatment 1 (respondents presented with personalized cost estimates for each type of fee based on their provided vehicle information), Voting Opportunity 2, Educational Treatment 2 (respondents watched an educational video developed for the purposes of this research discussing mileage fee privacy and mileage collection options as well as the equity / fairness of a gas tax compared to mileage fees and flat fees as is currently understood in the transportation funding / policy literature), Voting Opportunity 3, Reflection / Comment section, Demographics.
Respondents provided zip codes in the demographics section. These were spatially intersected with USDA RUCA codes to create a community-type variable. The RUCA codes were then aggregated to a smaller set of variables for modelling as shown below.
RUCA Code |
Description |
Aggregated RUCA Codes |
1 |
Metropolitan area core: primary flow within urbanized area |
Area core |
2 |
Metropolitan area high commuting: primary flow 30% or more to a UA |
High commuting |
3 |
Metropolitan area low commuting: primary flow 10% to 30% to a UA |
Rural |
4 |
Micropolitan area core: primary flow within an urban cluster of 10,000 to 49,999 (large UC) |
Area core |
5 |
Micropolitan area high commuting: primary flow 30% or more to a large UC |
High commuting |
6 |
Micropolitan area low commuting: primary flow 10% to 30% to a large UC |
Rural |
7 |
Small town core: primary flow within an urban cluster of 2,500 to 9,999 (UC) |
Area core |
8 |
Small town high commuting: primary flow 30% or more to a UC |
High commuting |
9 |
Small town low commuting: primary flow 10% to 30% to a UC |
Rural |
10 |
Rural areas: primary flow to a tract outside a UA or UC |
Rural |
Respondents provided responses to 15 questions about various attitudes and beliefs using a 5-point Likert scale. Common factor analysis with the primary axis method (a maximum likelihood approach) in the R psych package was used to create a reduced number of variables that capture a latent and broader set of attitudes and beliefs held by respondents. A parallel analysis scree plot was used to identify the number of factors and an orthogonal (varimax) rotation was used to develop final factor loadings. Factor scores were estimated for each respondent using the Thurston method (a regression approach) in the R psych package and used in our regression modeling.
For any additional questions, feel free to contact the researchers (Clare Nelson and Gregory Rowangould) at clare.nelson@uvm.edu or gregory.rowangould@uvm.edu.