Provisional LSOA populations method

Introduction

ONS provide mid-year estimates of population for lower layer super output areas (LSOAs). We need these in order to calculate IMD decile populations, as well as for small area analysis. Publication of the estimates is normally later than the equivalent mid-year estimates for lower tier local authorities (LTLAs), and has been subject to longer delays since the 2021 census.

In order to calculate provisional rates for publication before we have final versions of all the source data, a method has been developed by OHID to produce provisional LSOA populations. This page outlines the method, which can be applied to either published mid-year population estimates or population projections at LTLA level.

Terminology

The term ‘estimate’ is used for periods for which ONS have published ‘final’ mid-year estimates (although final, they can be revised). The term ‘projections’ is used for periods for which ONS have not yet published mid-year estimates.

Provisional figures calculated by OHID are provisional estimates if they are constrained to ONS mid-year estimates, and provisional projections if they are constrained to ONS projections.

Method

We combine the latest year (denoted \(Y_{base}\)) of published mid-year estimates for LSOAs with the projections or mid-year estimates for LTLAs for the year we need (\(Y_{target}\)). All calculations are carried out for males and females separately.

  1. Start with the single year of age (SYOA) LSOA populations for year \(Y_{base}\).

  2. ‘Age’ the populations by adding \(Y_{target}-Y_{base}\) to the ages.

  3. Replicate the \(Y_{base}\) age zero populations for \(Y_{target}\) age zero populations, and for other missing ages if \(Y_{target}-Y_{base}>1\).

  4. For the maximum age group (eg 90+), add the \(Y_{base}\) populations for all ages that reach that maximum age group in the target year (eg sum the age 89 and age 90+ \(Y_{base}\) populations to get the 90+ \(Y_{target}\) population if \(Y_{target}-Y_{base}=1\)).

  5. Aggregate these new aged SYOA populations within LTLAs.

  6. Divide each of the aged SYOA LSOA populations by the relevant aggregated aged SYOA LTLA estimate for \(Y_{target}\) and multiply by the equivalent ONS SYOA LTLA estimate/projection for \(Y_{target}\).

This gives provisional SYOA LSOA populations which sum to the ONS LTLA figures but reflect the individual LSOA age distributions as accurately as possible. They may also be aggregated to IMD deciles/quintiles.


Page last updated: May 2025