Data methodology

What data source is used?

House Price Watch uses HM Land Registry price paid data for England and Wales sold prices. Households and Population come from the Nomis Census 2021 postcode resident and household estimates (Office for National Statistics, Crown copyright). Boundary and map context is based on OpenStreetMap. Postcode centroid coordinates are sourced from the Office for National Statistics (ONS).

How are prices calculated?

Prices are calculated from recorded HM Land Registry sale transactions at postcode level, then aggregated to the selected location. We use these records to calculate average prices by month and year, identify the most expensive homes sold in each location, and generate additional summary statistics shown across the site. All figures reflect completed sales, not asking prices or valuations.

How many people and households are in each location?

Population is the Census 2021 count of usual residents (including people in communal establishments such as halls of residence) for every unit postcode that falls inside the location, summed from Nomis table P001. Households is the occupied-household count from Nomis table P002 using the same postcode list. A university ward can therefore show high Population density and only typical Household density.

Household density and Population density are those totals divided by the location’s polygon area in square miles. Each density is then labelled Very High, High, Typical, Low, or Very Low from national quintiles across all published locations (nested areas sit in the same pool). The raw totals are not labelled that way, because a large region will always outrank a small ward on headcount.

How is green space measured?

Green space comes from OS Open Greenspace: public parks and gardens, playing fields, play spaces, allotments, sports facilities, cemeteries, churchyards and golf courses. Each site's outline is laid over the location's polygon, and only the part inside the location counts towards its area. Sites that overlap (a play space inside a park) are counted once.

The green space total leaves out golf courses, because most need a membership; they still appear in the breakdown. It is not a measure of how green a place looks: farmland, most woodland and private gardens are not in OS Open Greenspace, so rural areas can show little public green space. Green space per resident divides the total by Census 2021 Population and is labelled Very High to Very Low from national quintiles, like density. We also say what share of areas have less green space per resident, and compare each area with its parent area and with the median (typical) area. To make square metres easier to picture, we convert them into full-size football pitches (105 m by 68 m, about 7,140 m²) per 1,000 residents. Park counts and named-site lists use a point inside each site, so a park on a border is listed in one area only.

Contains OS data © Crown copyright and database right 2026.

How is the data pipeline built?

The process is: (1) postcode centroids from ONS are matched and grouped into locations defined by OpenStreetMap boundaries and hierarchy; (2) HM Land Registry sale records for those postcodes are joined in; (3) Census 2021 Households and Population for those postcodes are summed, densities are calculated from polygon area, and national Density bands are assigned; (4) location-level outputs are generated, including average prices, properties sold, top sales, property type mix, price-range distributions, Households, and Population.

What does the location hierarchy mean?

Location pages are nested. For example, England and Wales data rolls down into regions and then smaller local areas. Rankings, charts, and summaries are generated using the selected level in that hierarchy. The location hierarchy and boundaries are sourced from OpenStreetMap (Open Street Map).

How often is data updated?

Each page includes a "Last updated" timestamp and the latest data coverage month. This indicates when the local dataset was last refreshed and up to which period Land Registry data has been included. Land Registry data is updated monthly, but it is usually delayed by about one month. Census 2021 Households, Population, and density figures are as at Census Day, 21 March 2021; they do not change when a new Price Paid Data file is imported.

What should this data be used for?

Use this data to understand long-term market trends, price ranges, relative differences between locations, and how densely a location is populated. It should not be treated as a current valuation of an individual property, and Census 2021 counts are not a live population estimate.