White British share: comparing local change with the national trend

Method correction, 14 October 2026. Earlier versions called the composition adjustment an age effect and the local residual a migration effect. The calculation uses ethnic shares alone. Those labels, the causal headline and the claim about what would happen without local migration were unsupported and have been removed. The figures below are descriptive components, not estimates of causes. The national component uses summed local authority rows, which can differ slightly from published national rows.

Earlier correction, 14 October 2026. Historical record; read with the method correction above. On 12 October this piece gave the national effect as -6.2pp, computed over the 307 authorities with a 2011 Census record of their own, and left out the eleven councils created after 2011. Each of those eleven is a merger of whole 2011 districts, so its 2011 figure is now the sum of those districts’ counts, and the decomposition covers all 318 authorities, which together make up England and Wales. The national effect is now -6.1pp: the White British share fell from 80.5% to 74.4% (Census 2011 KS201EW, Census 2021 TS021), the England and Wales figures. Each authority’s 2021 share is now taken from TS021 as published; for 259 authorities the site had used the model’s own 2021 base, a different ONS extract that differs from TS021 by up to 0.05 points. Burnley’s split is 80% national trend, 5% age structure and 15% local, against 81%, 6% and 13% on 12 October. The national trend is now the largest component in 275 of the 318 areas, against 266 of 307; Gwynedd, named below on 12 October as an area where the local component was the largest, is now led by the national trend.

Earlier correction, 12 October 2026. Historical record; read with the corrections above. This piece originally gave the national effect as -6.4pp, said the decomposition covered 305 local authorities, and said the national White British share fell from 80.1% to 74.5%, which is a fall of 5.6 points, not 6.4. The -6.4pp was not computed from the Census for 2011: every 2011 population total in the site’s data was zero, because the script that read the 2011 Census table did not recognise its “All usual residents” row, so the calculation fell back to a typed 80%. The data has held 307 decomposed areas throughout, and the 80.1% and 74.5% match neither the data nor the Census. With the real totals, the national effect is -6.2pp: across the 307 authorities whose boundaries allow a 2011 to 2021 comparison, the White British share fell from 79.7% to 73.5% (Census 2011 KS201EW and Census 2021 TS021). For England and Wales as a whole it fell from 80.5% to 74.4%. Burnley’s split is now 81% national trend, 6% age structure and 13% local, against 84%, 5% and 11% before. The piece also named Tower Hamlets and Newham as areas where local migration dominates; in this decomposition both are led by the national trend, before and after the correction.

Burnley’s White British share fell from 85.5% to 77.9% between 2011 and 2021. A descriptive calculation separates that change into:

• National trend: -6.1 percentage points, or 80% of Burnley’s change.
• Composition adjustment: -0.4 percentage points, or 5%.
• Local residual: -1.2 percentage points, or 15%.

Burnley’s total change is -7.6 percentage points. Components and the total are rounded separately, so the displayed components do not add exactly to the displayed total.

The national component is the change in White British share across the summed counts of all 318 local authorities. The composition adjustment multiplies the difference between an area’s starting share and the national starting share by the national proportional change. The residual is the local change left after subtracting those two components.

This calculation describes how local change compares with an aggregate benchmark. It uses no age or migration observations, so it cannot say how much of the change was caused by either, or what would have happened without migration.

The national component is the largest in absolute size in 275 of the 318 areas. The residual is largest in the other 43. It is negative in 28 of those and positive in 15. A positive residual offsets part of the decline represented by the other components. The place pages name the largest component in the direction of the overall change and identify a larger opposing component where present.

Data and method: ONS Census 2011 table KS201EW and Census 2021 table TS021, local authorities in England and Wales, via Nomis bulk downloads. For the eleven councils created after 2011, the 2011 count is the exact sum of the districts that formed them. The national component uses the sums of those local authority rows. These give 59,597,569 residents and 44,355,040 White British residents in 2021; the published England and Wales row gives 59,597,542 and 44,355,038 respectively. Census disclosure controls mean rows at different geographic levels need not add exactly. Both yield a White British share of 74.4% when rounded to one decimal place.

Percentages of Burnley’s change and counts of the largest component use unrounded components. Place chart labels use the displayed one-decimal components, so ties at that precision are named equally. This is the site’s descriptive share calculation, not an age-cohort decomposition.

Sources: ONS national ethnicity bulletin and ONS Census 2021 disclosure control methodology.