PaperlayerView exact revision on arXiv

Source-linked AI summary

Disadvantaged Communities Have Lower Access to Urban Infrastructure

Leonardo Nicoletti, Mikhail Sirenko, Trivik Verma

arXiv:2203.13784v1physics.soc-phcs.CYcs.SI

TL;DR

Unequal access to urban infrastructure is associated with broader inequalities among urban communities, making its spatial and demographic distribution important to understand. The paper uses an open-source framework to assess accessibility across cities and community profiles, finding a universal log-normal distribution and structural inequalities among similar community archetypes in the North American cities studied.

  • Problem

    Urban communities differ in access to services and amenities, and residents in areas with poor access tend to face disadvantages that can compound into social exclusion and reduced upward mobility.

    Source: Introduction · pp. 2–3

  • Method

    The authors use open-source data to assess accessibility to more than 50 types of amenities across 54 cities and examine socioeconomic community profiles in 10 cities.

    Source: Introduction · p. 4

  • Results

    Accessibility follows a universal log-normal distribution across 54 cities, and the 10-city assessment finds structural inequalities among similar urban-community archetypes across the North American cities studied.

    Source: Introduction · p. 4

  • Takeaways & Limitations

    The modular, reproducible framework can support municipalities in identifying regions and social groups facing access inequalities and adapting analyses to stakeholder values.

    Source: Discussion · p. 17

    The square-grid spatial units partially address the Modifiable Areal Unit Error but do not represent on-the-ground spatial contexts and introduce MAUP themselves.

    Source: Spatial Units · p. 26

Abstract

from arXiv · show

Disparity in spatial accessibility is strongly associated with growing inequalities among urban communities. Since improving levels of accessibility for certain communities can provide them with upward social mobility and address social exclusion and inequalities in cities, it is important to understand the nature and distribution of spatial accessibility among urban communities. To support decision-makers in achieving inclusion and fairness in policy interventions in cities, we present an open-source and data-driven framework to understand the spatial nature of accessibility to infrastructure among the different demographics. We find that accessibility to a wide range of infrastructure in any city (54 cities) converges to a Zipf's law, suggesting that inequalities also appear proportional to growth processes in these cities. Then, assessing spatial inequalities among the socioeconomically clustered urban profiles for 10 of those cities, we find urban communities are distinctly segregated along social and spatial lines. We find low accessibility scores for populations who have a larger share of minorities, earn less, and have a relatively lower number of individuals with a university degree. These findings suggest that the reproducible framework we propose may be instrumental in understanding processes leading to spatial inequalities and in supporting cities to devise targeted measures for addressing inequalities for certain underprivileged communities.

Introduction

The introduction frames unequal access to urban amenities as a spatial and socioeconomic problem, with consequences for residents and a need for targeted policy. It presents an open-source framework for examining accessibility across cities and demographic groups.

Introduction

  • Gentrification, housing-market dynamics, residential segregation, and exclusionary zoning have pushed low-income minority groups toward deteriorated areas receiving a smaller share of public investment.The introduction links these processes to concentrated urban poverty and inequalities among communities.Source: Introduction · p. 2
  • Residents in areas with few accessible services and amenities tend to face compounding disadvantages in education, health, job opportunities, social inclusion, and upward mobility.Source: Introduction · p. 3
  • Policies such as complete streets aim to improve access to services, shops, and recreation while supporting active lifestyles, public spaces, and economic vitality.The introduction argues that infrastructure can support shared spaces and civic life for underprivileged communities.Source: Introduction · p. 3
  • Systematically examining the spatial distribution of access alongside demographic distribution can inform targeted, equitable policies, while case-specific approaches may not address causes of urban inequality.Accessibility is defined as residents’ ease of reaching amenities or opportunities, and studies often assess access to specific infrastructure by demographic attributes.Source: Introduction · pp. 3–4
  • Across 54 cities, the framework finds accessibility follows a universal log-normal distribution; assessment of 10 cities identifies structural inequalities among similar urban-community archetypes across the North American cities studied.The framework uses open-source data and considers more than 50 types of amenities.Source: Introduction · p. 4

Results

Accessibility varies across cities and population groups: some cities have many residents in low-access areas, while disadvantaged profiles often have lower access than advantaged ones. In Chicago, the most disadvantaged group’s average accessibility was 27% lower than the least disadvantaged group’s.

There is more to the scale-free nature of accessibility 2

  • Across cities, accessibility scores follow normal distributions or mixtures of normal distributions, and scores decrease with the log of spatial-unit rank, with a cutoff for some regions.The authors suggest the cutoff may indicate regions that are not economically well integrated, depending on city sprawl.Source: There is more to the scale-free nature… · pp. 5–6, Figure 1 · p. 6
  • Around 90% of Houston and Seattle residents lived in spatial units with accessibility scores below 0.3.The population-based curves are compared with accessibility distributions across equal-sized spatial units.Source: There is more to the scale-free nature… · p. 7, Figure 2 · p. 8
  • The S(x) benchmark represents a hypothetical city where 50% of residents have accessibility of at least 0.5; no studied city matched it.A larger share of residents has higher accessibility when a city’s curve shifts right toward the benchmark.Source: There is more to the scale-free nature… · p. 7, Figure 2 · p. 8
  • The authors compare cities by population-curve error from S(x) and use quadrants to distinguish spatial access from population density patterns.The quadrants describe cases where access and population density are unevenly aligned, including high access in many spatial units but lower access where most residents live.Source: There is more to the scale-free nature… · p. 9, Figure 3 · p. 10
  • The analysis clusters households by socioeconomic attributes into three to five profiles per city; Chicago had five profiles.Attributes include ethnicity, income, household composition, language, education, employment, occupation, and commuting mode.Source: There is more to the scale-free nature… · pp. 11–12
  • 27% lower average accessibility separated Chicago’s most disadvantaged group from its least disadvantaged group.The least-accessible profile, excluding a suburban profile, had the lowest income, the largest share of visible minorities, the fewest university-educated residents, and the highest unemployment.Source: There is more to the scale-free nature… · p. 12
  • In a majority of the 10-city subset, low-income communities with larger minority shares and lower education levels typically had less access to urban amenities.In Chicago, the greatest category-level inequalities involved food choices, entertainment, active living, and health and well-being infrastructure.Source: There is more to the scale-free nature… · p. 13, Figure 4 · p. 14

Discussion

The discussion finds that infrastructure accessibility follows a universal log-normal pattern while socioeconomically disadvantaged groups remain underserved. It presents the framework as a basis for targeted investigation, while noting that findings depend on how accessibility is defined.

Discussion

  • The accessibility distribution follows a universal log-normal pattern, and the most socioeconomically disadvantaged groups are structurally underserved compared with the least disadvantaged.Gaps also appear in access to food, culture and entertainment, active living, and health infrastructure.Source: Discussion · pp. 16–17
  • The modular framework can help municipalities identify regions and social groups facing access inequalities, and can be adapted to stakeholder values and local policy goals.The authors suggest combining active-accessibility measures with dynamic opportunity measures for a broader view of poorly accessible neighborhoods.Source: Discussion · p. 17
  • Overall accessibility levels do not establish communities’ needs or contextual factors, so local case studies should add data on socioeconomic conditions over time and accessibility indicators.The framework provides an empirical starting point for decision-makers to investigate opportunities to address poor access.Source: Discussion · p. 18
  • Accessibility results depend on category weights shared across spatial units and cities; drastic weight changes can substantially alter distributions.The authors propose region-specific weights and sensitivity analysis to assess alternative definitions of accessibility.Source: Discussion · pp. 18–19
  • Despite equity-focused policies in the studied cities, accessibility gaps indicate that urban change processes and past or current policies still perpetuate social inequalities.The authors argue present policies must be more ambitious to reduce gaps and improve minimum access affordability.Source: Discussion · p. 19

Materials and Methods

The study measures accessibility through weighted walking distances to amenities and compares accessibility distributions across demographic groups. Its framework combines census clustering with network-based accessibility analysis to evaluate equity across cities.

Accessibility Score

  • Accessibility scores derive from pedestrian-network walking distances to seven amenity categories, weighted and aggregated, then log-transformed, normalized within each city, and inverted so higher scores indicate greater access.POIs are assigned to the nearest pedestrian-network edge, and distances follow weighted shortest paths rather than straight-line distance.Source: Accessibility Score · pp. 19–21

Data-driven Framework

  • The framework combines census-based social-group clustering, street-network and amenity analysis, and comparisons of accessibility distributions across groups; it is automated and documented as open-source code.Figure 6 presents the framework for evaluating equity in accessibility to urban infrastructure.Source: Data-driven Framework · p. 22, Figure 6 · p. 23
  • The framework measures accessibility in 54 cities and examines demographic distributions in a subset of 10 cities.The subset comprises New York City, Chicago, San Francisco, Los Angeles, Houston, Seattle, Miami, Toronto, Vancouver, and Montreal.Source: Data-driven Framework · p. 22

Data Description

The study draws on census, OpenStreetMap, and population-grid data, then uses consensus clustering to characterize socioeconomic groups. Its 250-meter grid reduces some spatial-unit issues but has limitations in data quality and boundary assignments.

Data Description

  • Census inputs include 46 variables on ethnicity, income, household and marital status, language, education, employment, occupation, and commuting; Canadian data are from 2016 and US estimates from 2017.The analysis also manually selected 44 socioeconomic and housing-related variables for clustering.Source: Data Description · p. 23, Clustering Method · p. 25
  • OpenStreetMap supplies pedestrian networks and amenity locations, but its completeness and quality vary; selected cities had more than 80% data availability, while some rural and developing areas may be incomplete.Amenity counts may be duplicated, though the study emphasizes locations rather than counts.Source: Data Description · pp. 23–24
  • Population density comes from the European Commission’s 2015 GHSL grid, with 250-by-250-meter cells matching the accessibility units; the data sources span different years.The authors assume population density does not change drastically over a few years and do not infer characteristics of individuals from group-level data.Source: Data Description · pp. 24–25

Clustering Method

  • Consensus clustering combines multiple algorithms and uses a consensus function to select robust cluster labels; k-modes measures dissimilarity by counting mismatches across categorical attributes.The implementation uses five clustering algorithms, seven distance functions, and 10 replications; preprocessing includes 80% winsorization and standardization.Source: Clustering Method · pp. 25–26
  • The authors could not compare k-modes with other consensus functions because true cluster labels are unknown, and propose expert validation as future work.They also note that differences between some cities’ clustering metric values are hardly distinguishable.Source: Clustering Method · p. 25

Spatial Units

  • Accessibility units are 0.0625-km² squares from the GHSL grid; this choice partly addresses, but also introduces, the modifiable areal unit problem.Population and accessibility data are interpolated to the grid, which the authors say can introduce error; fewer than 2% of boundary-overlapping squares receive cluster labels according to checking order.Source: Spatial Units · pp. 26–27
Loading 2203.13784v1…