How this map works.
The map itself stays sparse on purpose. This page holds the fuller picture: what the Vulnerability score measures, how need is combined with supply to show where to act, where the data comes from, and the honest limits of an open-data prototype.
What the Vulnerability score is
Every one of Ireland's 18,919 Census small areas gets a single Vulnerability score from 0 to 100, where 100 is the greatest need.
The score combines four themes: Carer share, Deprivation, Demand and Hard to reach. Each input is percentile-ranked across all small areas, the inputs are averaged within each theme, and the four themes are averaged into the final score. This follows the percentile-rank method used by the CDC/ATSDR Social Vulnerability Index and the OECD Handbook on Constructing Composite Indicators, so unlike variables (a percentage, a deprivation score, a rate) become comparable. The four themes count equally: equal weights, chosen deliberately.
An average can hide what drives it. A score of 70 can mean moderately stretched on every front, or extreme on one. So each area also carries a flag count — how many of its four themes sit in the worst quarter of Irish areas — shown in the area's tooltip on the map. Three or four flags means severe across the board; one outlier theme cannot produce that. (This is the SVI's flag mechanism, the standard companion to an averaged score.)
The four themes
Each theme captures one pressure on carers. Together they make up the Vulnerability score.
Targeting measures: need meets supply
The Vulnerability score is about need alone. The two measures below add a second ingredient, the supply of in-home support, to show where need and a lack of nearby help line up. They are not drivers of vulnerability: they combine need with supply to point to where action would do the most good.
Support gap
Need × distance to nearest supportAn area scores high only when it has both high need and no support for carers nearby. The two are multiplied, not averaged, so one alone is not enough. "Support" here means home-care agencies and Family Carers Ireland centres. Distance is measured to the nearest office, following the USDA Food Access Research Atlas, the standard method for mapping service deserts.
Need vs support
A 3×3 need-by-supply viewThe same two ingredients shown side by side rather than fused: need on one axis, distance to support on the other, split into thirds. The dark corner is high need and far from help. It lets you see which axis drives each area, separating "high need but served" from "low need but remote", which a single score blurs. The two-axis grid is a Stevens bivariate choropleth, a standard way to show two measures at once.
The Nursing-home gap applies the same idea for nursing-home operators rather than in-home support: Demand (the older and disabled population, who are likely to need residential care) multiplied by distance to the nearest nursing home. It points to where that demand sits far from a home. It uses Demand, not the full carer Vulnerability, because residential care is driven by who needs care, not by how many informal carers an area has.
The HSE market-gap view
Home-support providers apply to serve one of the 20 HSE Integrated Healthcare Areas. So the map also rolls everything up to those 20 areas. For each, we take the population-weighted average of small-area need, count the approved providers serving the area, and subtract the supply rank from the need rank. A higher score means a wider gap between need and provision.
Subtracting ranks keeps the score bounded and symmetric: it runs from −100 to +100, where zero means an area's supply sits at the same rank as its need, and a negative score means the area is better supplied than its need would predict.
Care-service markers
Three point layers show where support for carers and older people physically exists: nursing homes (543, from the HIQA register), home-care agencies (206 distinct offices, from the 20 HSE approved-provider lists) and Family Carers Ireland centres (16). Addresses were geocoded with the keyless OpenStreetMap Nominatim service.
Geocoding precision is uneven. Eircodes are not openly geocodable, so home-care offices are the weak layer: only about 27 of 201 resolved to a street, most landed at the town centre, and some at the county. Every point carries a precision flag, and the error matters least in rural deserts where true distances are tens of kilometres.
Nursing homes are residential care, not in-home help. They make up roughly 71 percent of the dots, but a nursing home is a place someone moves into; it replaces home caring rather than supporting the informal carer. They stay visible for context but are excluded from the Service desert score.
Data sources
Everything is keyless open data for the Republic of Ireland, verified live against the source files on 8 June 2026.
Limitations
These are the known edges of a keyless, open-data prototype. The rule of thumb: everything here is an area-level signal to investigate, never a verdict about a place or a person.
Area-level rates, not individuals. The map shows where risk factors cluster by area. Small-area census data cannot identify any individual carer.
Scores are 0 to 100 ranks, not percentages. They place an area against the rest of Ireland, so colours are relative, not absolute thresholds.
Deprivation is borrowed from the larger area. The Pobal score is not openly released at small-area level, so each small area inherits its parent Electoral Division's value. Big differences between areas survive; fine variation within a division is lost.
Disability is the total rate, not severity. The census gives only the overall disability figure at small-area level, used here as a reasonable stand-in.
Geocoding is approximate. Offices are placed by reading their address text, often only to the town. Distances are straight-line, so real journeys, especially in the rural west, are longer than the map implies.
Provider size is unknown. The data shows where providers are, not their capacity or how busy they already are, so a small agency and a large one count the same.
Averages hide spread. A region or HSE-area colour is the average of its small areas, so a calm-looking area can still hide pockets of severe need. Zoom in before concluding.
It is a snapshot. Built on Census 2022 and provider registers as of mid-2026. Registers change as providers open and close; the census runs about every five years.
Republic of Ireland only. Northern Ireland uses separate census geography (NISRA) and a different deprivation index (NIMDM). An all-island version is a distinct piece of work.