Comparing the Promise and Reality of E-Scooters: A Critical Assessment of Equity Improvements and Mode-Shift

0.00

Add to Wishlist
Add to Wishlist

Description

Comparing the Promise and Reality of E-Scooters: A Critical Assessment of Equity Improvements and Mode-Shift, Is A Well-Researched Topic, It Is To Be Used As A Guide Or Framework For Your Research.

Abstract

In just three years, e-scooters have substantially disrupted and altered the urban mobility landscape. Throughout this period, they have been commonly touted as part of a larger micromobility solution that promises to erase equity barriers and solve the first-mile/last-mile problem. However, few studies in the nascent e-scooter literature have considered these claims. In this study, we surveyed students at Portland State University (n = 1,968) about the role that e-scooters, among other modes, played in meeting their general and university-related travel needs. We then estimated models that incorporated demographics, travel behavior, and latent attitudes distilled using exploratory factor analysis (EFA). These models were used to assess the current performance of e-scooters in meeting equity and mode-shift goals.
We first estimated ordinal logit models to understand the relationship of these factors to the stated number of trips taken in the 7 days prior to the survey by e-scooter, car, bike, and MAX light rail. Perceived propensity to switch to using e-scooter, car, bike, or MAX light rail modes for commuting to the university should their present primary commute mode became unavailable. We also designed and implemented a stated choice experiment (SCE) consisting of several hypothetical scenarios of a commute to PSU. In the SCE, students were given a three-mode labelled set consisting of car, bike, and e-scooter + MAX choices. The experiment choice sets were designed using a D-Efficient method. In order to understand the relationship of travel time and cost in addition to the other covariates on mode choice, we estimated a multinomial logit (MNL) model from the experiment data. We used this model to perform a thorough sensitivity analysis to uncover the most impactful factors that encourage first-mile/last-mile e-scooter usage. Additionally, we used the model to generate catchment area maps for e-scooter/MAX multimodal trips in the Portland area based on the most probable mode choice at every point in the city.

In addition to the models, we asked students to indicate barriers that actively prevented them from using non-auto modes more often. We mapped out barrier “hot-spots” in order to understand the current travel realities of the city.

Results were mixed in indicating that e-scooters bring about racial and gender equity in transportation. Additionally, we found that there was no place in the city where taking an e-scooter to connect with MAX to travel to PSU was more preferable (utilitarian) than taking a bike or private car, on average. This suggests that e-scooters are currently not a practical solution to the first-mile/last-mile problem. However, our findings revealed “dials” that can be tweaked through policy measures in order to promote this kind of use. Overall, our critical analysis of the implementation of e-scooters suggests that their promise is overstated, at least without substantial policy changes to encourage desired use cases.

Table of Contents

Abstract ……………………………………………………………………………………………………………….. i
Dedication …………………………………………………………………………………………………………… iii
Acknowledgements ………………………………………………………………………………………………. iv
List of Tables ……………………………………………………………………………………………………… viii
List of Figures ………………………………………………………………………………………………………. ix
1 Introduction ………………………………………………………………………………………………….. 1
2 Literature Review ………………………………………………………………………………………….. 6
2.1 E-Scooters: Potential and Performance ……………………………………………………… 6
2.2 Approach and Contribution…………………………………………………………………….. 10
3 Method ………………………………………………………………………………………………………. 12
3.1 Survey Design ……………………………………………………………………………………….. 12
3.1.1 Experimental Design ……………………………………………………………………….. 12
3.1.2 Levels …………………………………………………………………………………………….. 18
3.1.3 Mode Attitudes and Decision Factors ………………………………………………… 22
3.2 Administering the Survey ……………………………………………………………………….. 25
3.2.1 Qualtrics Implementation ………………………………………………………………… 27
3.2.2 Incentive ………………………………………………………………………………………… 29
3.2.3 Outreach ……………………………………………………………………………………….. 30
3.3 Data Cleaning ……………………………………………………………………………………….. 33
3.4 Models …………………………………………………………………………………………………. 35
3.5 Spatial Analyses …………………………………………………………………………………….. 35
3.5.1 Spatial Distribution of Barriers………………………………………………………….. 35
3.5.2 Spatial Implementation of MNL Results …………………………………………….. 38
4 Results ………………………………………………………………………………………………………… 43
4.1 Descriptive Statistics ……………………………………………………………………………… 43
4.2 Attitude Indices …………………………………………………………………………………….. 50
4.2.1 Reliability ……………………………………………………………………………………….. 50
4.2.2 Correlations ……………………………………………………………………………………. 52
4.2.3 Indices Calculated Using Item Sums and Means …………………………………. 54
4.2.4 Principal Components Analysis (PCA) and Exploratory Factor Analysis (EFA) 54
4.2.5 Calculated Latent Variable Scores …………………………………………………….. 65
4.2.6 Structural Equation Modelling (SEM) ………………………………………………… 66
4.2.7 Index method choice ………………………………………………………………………. 68
4.3 Number of Trips Ordinal Regression ………………………………………………………… 69
4.4 Propensity to Switch Linear Regression ……………………………………………………. 84

4.5 Stated Choice Experiment MNL Regression …………………………………………….. 100
4.6 Spatial Distribution of Mode Use Barriers ………………………………………………. 108
4.6.1 Individual Mode Findings ……………………………………………………………….. 108
4.6.2 Comparison of Similar Barriers Across Modes ………………………………….. 113
4.7 MNL Sensitivity Analysis and Visualization ……………………………………………… 119
4.7.1 Experiment Attributes …………………………………………………………………… 122
4.7.2 Demographics ………………………………………………………………………………. 129
4.7.3 Latent Attitudes ……………………………………………………………………………. 135
4.7.4 Example of obtaining “good” first-mile ridership ………………………………. 142
5 Discussion and Conclusion…………………………………………………………………………… 145
5.1 Barriers ………………………………………………………………………………………………. 145
5.2 General Model Findings – Travel behavior, routine, built environment, and attitudes ………………………………………………………………………………………………………. 146
5.3 E-Scooter Specific Findings ……………………………………………………………………. 150
5.4 Policy Implications ……………………………………………………………………………….. 153
5.5 Limitations ………………………………………………………………………………………….. 155
5.6 Future Research ………………………………………………………………………………….. 157
5.7 Conclusion ………………………………………………………………………………………….. 158
References ………………………………………………………………………………………………………. 161
Appendices ………………………………………………………………………………………………………. 165
A. Pearson Correlations of Scale Items ………………………………………………………. 166
B. Full Latent Variable Estimation Method Comparisons ……………………………… 172
C. Full OLS Models …………………………………………………………………………………… 187
D. Full Logit Models …………………………………………………………………………………. 199
E. Full MNL Models …………………………………………………………………………………. 216
F. Survey Instrument ……………………………………………………………………………….. 225

Additional information

Author

Michael Glenn McQueen

No of Chapters

5

No of Pages

255

Reference

YES

Format

PDF

Reviews

There are no reviews yet.

Only logged in customers who have purchased this product may leave a review.