Technical note
How the unpaid-care map was calculated.
This note explains the data, geography and assumptions behind the Hidden Week map. The central idea is simple: take the unpaid-care hours people reported in Census 2021, count them cautiously, and show how that hidden labour is distributed across local neighbourhoods.
Who is counted?
The main data source is Census 2021 table TS039, provision of unpaid care. It counts usual residents aged 5 and over by the number of hours of unpaid care they reported providing each week. The figures refer to Census Day, 21 March 2021.
In plain terms, this is about people who provide unpaid help or support because someone has long-term physical or mental health conditions, disability, or problems related to old age. It does not count children aged under 5, because they were not asked the unpaid-care question.
The table identifies the person providing care, not the age of the person receiving it. For that reason, the public story does not describe these figures as care for people aged 65 and over only. A 65+ recipient-specific estimate would need another source or a clearly labelled modelling assumption.
How are care-hours estimated?
Census 2021 gives unpaid care in bands rather than exact hours. To avoid exaggerating, the map uses a lower-bound estimate:
- 1 to 9 hours is counted as 1 hour.
- 10 to 19 hours is counted as 10 hours.
- 20 to 34 hours is counted as 20 hours.
- 35 to 49 hours is counted as 35 hours.
- 50 or more hours is counted as 50 hours.
The result is a conservative minimum number of unpaid care-hours per week. Many people in each band will provide more than the lower-bound value used here, especially in the 50+ hours category.
What does the map height show?
For each lower layer super output area, the map calculates:
Minimum unpaid care-hours per week ÷ residents aged 5 and over × 100.
This gives a comparable local burden: minimum unpaid care-hours per 100 residents aged 5 and over each week. The total across the five Tyne and Wear local authorities is at least 2,648,586 unpaid care-hours per week.
Why not age-standardise this?
The map is designed to show the actual burden of unpaid care being carried in local places, not the underlying age-adjusted propensity to provide care. Age structure matters, and some neighbourhoods will have more unpaid care partly because they have more people at ages where care needs are higher.
That is not a flaw for this purpose. If a place carries more care because it has more older or disabled residents, that is still real work being carried by real people. A different analysis could age-standardise care rates to ask a different question.
Deprivation, health and disability
Deprivation is shown using the English Indices of Deprivation 2025, specifically the Index of Multiple Deprivation decile. D1 means the most deprived 10% of neighbourhoods in England, and D10 means the least deprived 10%.
Bad or very bad health comes from Census 2021 table TS037. Disability limited a lot comes from Census 2021 table TS038. These are shown to identify places where unpaid care, poor health, disability and deprivation stack together.
Geography
The geography is lower layer super output areas in Newcastle upon Tyne, Gateshead, North Tyneside, South Tyneside and Sunderland. These are small statistical neighbourhoods, not named communities. For the public map, LSOA counts are aggregated to 2024 wards using the official best-fit lookup. The circles therefore show named ward-level burden, while the background keeps the small-area pattern visible.
Limitations
These are self-reported Census data from 2021. The figures do not show whether care is adequate, whether carers are supported, or whether the person receiving care lives in the same household. The care-hours total is deliberately conservative and should be read as a lower-bound estimate.
The figures also do not identify the age of the person receiving care. They should therefore be read as unpaid care provided by residents aged 5 and over, not as unpaid care for older people aged 65 and over.
The deprivation data are from IMD 2025, while the care, health and disability data are from Census 2021. The deprivation overlay should be read as area context, not as proof of causation.