The goal of this study is a definitive understanding of the differences in attitudes and behaviors among the US population with respect to public transportation and neighborhood choice. We aim to understand which characteristics and beliefs are behind those differences. To that end, we conducted a large online survey (11,842 respondents) across 46 Metropolitan Statistical Areas (MSAs) in the United States. The selected MSAs span the full geography of the U.S. and include some cities with well-developed transit systems and others with less developed transit system. The sample also ensured minimum quotas for all age groups, allowing the study to compare different generations, geographies, and neighborhood types.
The results reveal that the most important factors in determining whether someone is at least an occasional transit user are:
- High population density of home neighborhood (positive effect)
- Being employed or a student (positive effect)
- Being an ethnic minority (positive effect)
- High-quality local transit (positive effect)
- High income (negative effect)
Surprisingly, education level and the presence of children in the home do not appear to have a strong association with transit use either way when the other variables are controlled for. This suggests that despite high rates of transit use in college, most former students do not continue to ride transit after that experience. People with kids, meanwhile, may be just as willing as others to take transit when it is available in their neighborhoods.
We are able to explore what factors generally draw people to public transportation. Travel time, reliability, and cost appear to be more important than “flashy” features like Wi-Fi. Additionally, people who are offered pretax transit commuter benefits by their employers are over five times as likely to take transit regularly as employed persons who are not receiving benefits.
The large sample size allows for comparisons across geography, age group, quality of local transit, levels of transit use, levels of population density, and other characteristics. We see the most variation across age groups. Behavior changes considerably along the age spectrum, even when controlling for other factors such as employment, household income, and neighborhood type.