Skip to main content
← Back to Methodology

Methodology

Indices and features

The Service Delivery Index: what it measures

Each suburb with household survey coverage displays a Service Delivery Index from 0 to 100. The index is a composite of four infrastructure access rates drawn from the City of Cape Town 2024 Household Survey: piped water in the dwelling, flush toilet connected to sewerage, weekly waste removal, and grid electricity access.

// Service Delivery Index: equal-weighted composite of four access rates

score = (piped_water_pct + flush_toilet_pct + waste_removal_pct + grid_electricity_pct) / 4

// Each component = % of surveyed households with access to that service

// Index range: 0 (no services) to 100 (universal access across all four dimensions)

// City median computed at build time across all 407 surveyed suburbs

An index of 100 does not mean the suburb has no infrastructure problems, it means that in the survey period, all four basic services were reported as accessible by effectively all surveyed households. Outage frequency, water pressure, and collection reliability are not captured in this index.

South African context

The spatial legacy of apartheid-era planning means that service access is not evenly distributed across Cape Town. Suburbs in the Cape Flats and former township areas frequently show lower service delivery indices, not as a reflection of community behaviour, but as a direct consequence of historical under-investment in infrastructure. Where a suburb scores below the city median, StreetSignal frames this as a structural infrastructure gap, not a neighbourhood deficit.

Development Focus Areas

The City of Cape Town's District Spatial Development Framework (DSDF, January 2023) designates 27 Development Focus Areas across the metro. These are areas where the City has committed to prioritising public investment in infrastructure, services, and economic opportunity.

StreetSignal identifies suburbs that fall within a Development Focus Area using spatial intersection of DFA polygons against suburb boundary polygons. 43 of 744 suburbs fall within a designated Development Focus Area, including areas in Khayelitsha, Mitchells Plain, Salt River/Woodstock, District Six, Athlone, and Philippi.

This is an asset-framing indicator: it signals where the City is directing investment, particularly in historically under-resourced areas. Development focus areas indicate City planning intent as documented in the 2023 DSDF. Actual development timelines are subject to Council budget allocation and may change.

Source: City of Cape Town District Spatial Development Framework (January 2023), CC-BY open data.

New Development Areas

The City of Cape Town's District Spatial Development Framework (DSDF, January 2023) designates 836 individual land parcels as New Development Areas across the metro, classified by intended land use: High Density Residential, Medium Density Residential, Low Density Residential, Mixed Use, Industrial, Commercial, and Public Service.

StreetSignal assigns each NDA parcel to a suburb by testing the parcel centroid against suburb boundary polygons. 244 of 744 suburbs contain at least one NDA parcel. For each suburb, the total hectares and type breakdown are displayed.

New Development Areas indicate planning designation, not construction timelines. Not all designated areas will proceed to development. NDA parcels complement Development Focus Areas: DFAs signal strategic City investment priorities, while NDAs identify specific land parcels earmarked for particular uses.

Source: City of Cape Town District Spatial Development Framework (January 2023), CC-BY open data.

Free basic electricity

Free basic electricity is a constitutional entitlement (Section 27 of the Constitution) operationalised through the City of Cape Town's indigent support programme. Qualifying households receive a monthly allocation of free electricity via the Lifeline tariff.

StreetSignal reports the proportion of prepaid electricity connections receiving free basic electricity in each suburb. This figure is calculated from the most recent 3 months of the City's Prepaid Electricity by Suburbs and Tariffs dataset. Suburbs with fewer than 300 total connections in the period are excluded as statistically unreliable.

Coverage: 357 of 744 suburbs. Coverage is incomplete for some areas where suburb naming in the electricity dataset does not match the platform's suburb index. This disproportionately affects townships where the City's prepaid system uses sub-area codes, block numbers, or address-level identifiers rather than suburb names. Khayelitsha, parts of Mitchells Plain, and several other large townships have insufficient matched connections to produce a reliable figure. The city-wide median is 5.6%.

The proportion reflects the reach of the City's indigent support in each suburb, not a measure of poverty. This figure covers prepaid connections only. Credit-metered households on the indigent register are not included.

Source: City of Cape Town Prepaid Electricity by Suburbs and Tariffs (January 2024 to February 2026), CC-BY open data.

Neighbourhood narratives

Each suburb page includes a structured narrative summary that describes the suburb across five paragraphs: overview, safety, property, services, and education. These narratives are template-based and generated at build time from the suburb's raw data files. They are not produced by a large language model or any generative AI system.

Every sentence in a narrative traces to a specific data point from the suburb's record. No values are interpolated, estimated, or inferred. Numeric values are inserted directly from the computed data layer. Safety descriptions use the canonical risk labels ("Lower reported crime", "Moderate reported crime", or "Higher reported crime") and always include the relative-index caveat.

Quality rules

  • British English throughout
  • No superlatives or competitive claims
  • No interpolated or rounded numbers beyond what the source data provides
  • Every sentence is independently verifiable against the suburb's data record
  • Paragraphs are omitted where the suburb lacks data for that dimension, rather than generating placeholder text

Why templates, not AI-generated prose

Template-based narratives guarantee deterministic output. The same data always produces the same text. This eliminates the risk of hallucinated statistics, inconsistent framing across suburb pages, or tone variation that could imply editorial bias. Every narrative can be audited by comparing the rendered text against the suburb's JSON record.

Comparison pages

StreetSignal generates pairwise comparison pages that allow users to evaluate two suburbs side by side. Pairs are selected by geographic proximity: for each of the 744 suburbs, the three nearest neighbours are identified using haversine distance from suburb centroids. This produces approximately 1,400 unique comparison pairs.

Pair selection and deduplication

  • Three nearest neighbours per suburb, computed by haversine distance between centroids
  • Alphabetical slug ordering for deduplication (e.g. bellville-vs-parow, never parow-vs-bellville)
  • Pairs where both suburbs lack safety data are excluded

Metrics compared

  • Safety score and risk profile
  • Property valuations (median value, price per m², CAGR)
  • School count and learner-educator ratio
  • Healthcare facility proximity
  • Civic responsiveness (median resolution time, same-day resolution rate)
  • Transport connectivity

Language conventions

Winner indicators use relative language throughout. A suburb with a higher safety score is described as having a "higher safety score", not as being "safer". A suburb with a lower median property value is described as "lower median property value", not "more affordable". This prevents the comparison from implying qualitative judgements that the data does not support.

Change detection

StreetSignal tracks how suburb metrics change over time by comparing snapshots of the data taken before each data update. Metrics are snapshotted before each update as a manual process, and trends are computed by comparing the two most recent snapshots. Trend indicators appear on suburb pages only after the second data update for a given metric.

Direction thresholds

A change must exceed a minimum threshold to be classified as a directional move. Changes within the threshold are classified as stable or unchanged.

  • Safety score: +/- 3 points
  • Property values: +/- 2%
  • School count, healthcare facility count: any change

Direction labels

  • Safety score, civic responsiveness: improved / declined / stable
  • Property values: increased / decreased / stable
  • School count, healthcare facility count: increased / decreased / unchanged

When trends are not shown

Trend indicators require at least two snapshots to compute a direction. For metrics where only one snapshot exists, no trend is displayed. This prevents the platform from implying stability or change where no comparison data is available.