Transport Poverty in the Amsterdam Metropolitan Area: Relationships with Socioeconomics and the Built Environment at the Neighborhood Level

Abstract

Limiting transport poverty is consequential in improving well-being and employment levels, which play meaningful roles in deciding public policy. We analyze how different socioeconomic and built environment factors are related to the transport poverty environment for car and public transport in terms of strength, significance, and direction for neighborhood zones within the Amsterdam Metropolitan Area—our study area. Additionally, we provide policy recommendations for the study area.

To this end, we use spatial distributions of environmental transport poverty indicators and perform weighted least-squares regression analyses, where we regress each transport poverty indicator on all built environment and socioeconomic variables. Our regression analyses are preceded by a combination of logarithmic variable transformation, insignificant variable elimination, and data normalization.

The environmental transport poverty indicators consist of average travel times, average single-trip travel costs, and the number of accessible jobs within thirty minutes. Regarding the built environment; we include inhabitant density and job density, whereas the socioeconomic characteristics in our analyses consist of household size, five age cohorts (0-18, 18-34, 35-54, 55-64, and $65^{+}$), gender, car ownership, and income.

Our results indicate that levels of environmental car transport poverty are fairly low over the whole study area when compared to the public transport poverty environment—the highest car transport poverty levels among all zones correspond with the lowest levels of public transport poverty. Regression results demonstrate that differences in the transport poverty environment are substantially correlated with differences in the zonal built environment and socioeconomic characteristics only for public transport travel times and public transport job accessibility, which are also the only transport poverty indicators to exhibit considerable variation among the study area zones in general.

Furthermore, the strength and significance of job density almost invariably greatly exceed those of inhabitant density and the socioeconomic variables—in relation to the transport poverty environment indicators. Inhabitant density is, despite being overshadowed by job density, also deemed to exhibit a marked correlation to the public transport measures of travel time and job accessibility. Both built environment variables are related to favorable transportation conditions. Remarkably, all socioeconomic variables display either insignificant or rather weak correlations to the transport poverty environment.

Consequently, we suggest focusing mainly on public transport measurements of travel times and job accessibility when analyzing differences in the transport poverty environment at a neighborhood level and the roles of job density and (to a lesser extent) inhabitant density herein. We visualize the combined spatial distributions of both job density and average public transport travel times, and inhabitant density and average public transport travel times for our study area. Areas potentially interesting for public transport improvement are identified based on the simultaneous occurrence of both relatively high levels of job or inhabitant density and relatively high travel times.