Liveability Index for Greater London

• Liveability Index for Greater London is a composite measure that evaluates neighbourhood quality through diversity, proximity, and population density of services.
• It employs a systematic methodology integrating urban service diversity, pedestrian-accessible proximity, and density using normalized, data-driven indicators compliant with OECD guidelines.
• Empirical analysis links higher liveability scores to increased pedestrian activity, offering actionable insights for urban planning and policy interventions.

The Liveability Index for Greater London is a composite assessment tool that captures the spatial variation, constituent drivers, and urban impacts of neighbourhood liveability, following both the principles of the 15-minute city paradigm and technical guidelines for composite urban indicators. It operationalizes liveability as the capacity of a neighbourhood to act as an urban attractor—facilitating access to services, fostering social interaction, and supporting active travel behavior. The Index provides a systematic and decomposable framework to identify strengths and deficiencies across London’s fine-grained neighbourhoods, directly linking features such as diversity, proximity, and density of urban services to pedestrian activity and broader urban vibrancy (Jeong et al., 10 Oct 2025).

1. Index Definition and Principal Domains

The Liveability Index for Greater London is explicitly constructed as a composite measure at the neighbourhood scale, evaluating the ability of a locality to concentrate daily services and urban functions. Three core domains structure the Index:

• Diversity: Quantified using Shannon’s entropy, this domain measures the variety of Points of Interest (POIs) classified into five urban service categories: commerce, education, entertainment, living, and healthcare. The diversity metric for living amenities, for example, is computed as $D_{lk} = -\sum_j P_{kj} \ln(P_{kj})$, where $P_{kj}$ is the probability of observing amenity type $j$ in neighbourhood $k$.
• Proximity: Measured via a hybrid methodology combining average Euclidean distances from a neighbourhood centroid to the center of mass of each service type, with network-based distances supplementing where necessary. This indicator reflects pedestrian-accessible reach.
• Population Density: Determined using LSOA-level population counts reprojected to neighbourhood boundaries and expressed per square kilometre. Sufficient density is essential for the viability and sustained operation of diverse services.

Collectively, these domains reflect the essential tenets of the 15-minute city paradigm—namely, the co-location, accessibility, and sufficiency of urban amenities that allow residents to satisfy regular needs locally.

2. Methodological Framework and Data Sources

The Index adheres to the OECD’s 10-step technical guidelines for composite indicators, ensuring a transparent and replicable construction:

• Indicators are calculated, then normalized via a rank-based exponential transformation (to support aggregate compensability and comparability), yielding values rescaled to $[0, 100]$.
• Principal Component Analysis (PCA) is used to examine the dimensionality and derive weights, though ultimately equal weights are adopted following empirical assessment and to ensure representation of orthogonal aspects.
• Robustness is tested through Monte Carlo simulations and sensitivity analyses (using Dirichlet distributions) to evaluate uncertainties.

Datasets deployed include:

• POIs from OpenStreetMap (OSM), assigned to urban service categories.
• The OSM street network (via OSMnx) for proximity calculations.
• February 2023 mobility data from Locomizer, providing normalized footfall scores at a fine hexagonal scale (H3 level 10, $\sim$76 m).
• Population statistics projected from LSOA granular census data.

This multimodal data integration allows for precise measurement of both neighbourhood features and the resulting urban activity.

3. Spatial Distribution and Index Decomposition

The Liveability Index reveals marked spatial heterogeneity across Greater London:

• Inner London: Contains the preponderance of high-liveability neighbourhoods, attributable to both higher diversity and better proximity to amenities, underlain by substantial population density. Boroughs such as Islington, Kensington and Chelsea, and Westminster consistently score highest—primarily due to strong diversity across commerce, education, and healthcare.
• Outer London: Generally records lower scores, though select peripheral neighbourhoods (often regional service hubs like major shopping centres and hospitals) maintain high diversity despite lower density and greater service distances.
• Domain Contributions: In top-ranked boroughs, the diversity domain dominates the composite score, while mid-ranked regions see increasing influence of proximity.

The Index is fully decomposable: analysis at the borough or neighbourhood level exposes which individual domains (e.g., entertainment diversity, healthcare proximity) drive scores and where strategic improvements are needed.

4. Association with Active Travel Behavior

A fundamental aspect of the Index is its empirical linkage to pedestrian activity:

• Ordinary Least Squares (OLS) Regression: Modelled as $Foot_k = \beta_0 + \beta_1 LI_k +\epsilon$, with $\beta_1 = 0.0084$ ($p < 0.001$), indicating each unit increase in liveability predicts a corresponding increase in footfall. A neighbourhood at the 75th percentile of liveability demonstrates $\sim$42% higher mean footfall than one at the 25th percentile.
• Geographically Weighted Regression (GWR): $$Foot_k = \beta_0(u_k, v_k) + \beta_1(u_k, v_k) LI_k + \epsilon_k$$, with $\beta_1$ ranging spatially from near zero to $0.0249$. Notably, marginal increases in liveability post greater footfall benefits in some peripheral neighbourhoods. The adjusted $R^2$ of 0.78 across local models affirms a strong relationship.

Both regressions confirm that liveable, amenity-rich, and accessible neighbourhoods act as urban attractors, directly boosting pedestrian flows and active travel.

5. Implications for Local Planning and Urban Policy

The detailed spatial patterning and decomposability of the Liveability Index have direct implications:

• Targeted Interventions: Policymakers can identify neighbourhood or borough-specific deficiencies (e.g., low proximity scores in outer boroughs) and prioritize investments such as improved transport links or amenity placement.
• Monitoring and Diagnostics: The Index supports ongoing monitoring, allowing local authorities to track the effects of interventions in service diversity, proximity, and density.
• Strategic Facility Allocation: Insights into local versus regional service provision guide decisions on where to invest in facilities that maximize both local vibrancy and regional accessibility.

The Index thereby informs strategies to foster more walkable, attractive, and resilient urban environments throughout Greater London.

6. Recommendations and Future Directions

Recommendations for enhancing neighbourhood liveability involve:

• Investing in diverse, high-quality amenities responsive to both local and regional needs.
• Leveraging the Index decomposition in planning documents to address the revealed weaknesses (e.g., augmenting healthcare diversity or improving entertainment proximity).
• Employing the Index as a tool for both benchmarking and temporal evaluation.

Research extensions proposed include:

• Incorporating temporal models (e.g., Geographically and Temporally Weighted Regression, GTWR) to capture seasonality or daily variations in footfall and liveability.
• Expanding indicator sets to include qualitative and socio-economic variables, enhancing concept validity.
• Longitudinal studies to examine causality in the relationship between improvements in liveability and changes in active travel.
• Comparing and integrating with alternative liveability constructs for transferability and generalization.

7. Mathematical and Statistical Formulations

The principal models underpinning the Index and its behavioural associations include:

• Shannon’s Entropy for Diversity:

$D_{lk} = -\sum_j P_{kj} \ln(P_{kj})$

• OLS Model for Footfall:

$Foot_k = \beta_0 + \beta_1 LI_k + \epsilon$

• GWR Model for Footfall:

$$Foot_k = \beta_0(u_k, v_k) + \beta_1(u_k, v_k) LI_k + \epsilon_k$$

• Global Moran’s I (spatial autocorrelation of residuals):

$$I = \frac{N}{W} \frac{\sum_i \sum_j w_{ij} (x_i - \bar{x})(x_j - \bar{x})}{\sum_i (x_i - \bar{x})^2}$$

These formulations enable robust inferential analysis, spatial mapping, and technical comparison.

---

The Liveability Index for Greater London thus provides an evidence-based, systematically constructed framework for measuring, diagnosing, and enhancing neighbourhood quality through the assessment of diversity, proximity, and density, underpinned by rigorous data and mathematical models. Empirical validation connects liveability directly to pedestrian activity, supporting its use in urban planning and policy to foster more attractive, equitable, and sustainable city environments (Jeong et al., 10 Oct 2025).