Impact of revisions to the mid-year population estimates

Revisions to the standard mid-year population estimates have a significant impact on some local authority resident populations, particularly in London and other large cities. The effect on calculated population data such as benefit claimant rates is also significant, and could lead to changes in how initiatives to tackle worklessness and other issues are targeted

Which local authorities are most affected?

Revised mid-year estimates for local authority resident populations were published in summer 2010, with revised small area estimates published in September. The overall national population has not been changed, only the way that this is split between local authorities and small areas.

Table 1 shows the 10 local authorities with the biggest increases in overall population under the revised population estimates. Seven of the ten LAs with the largest increases are in London, with the remaining three also in large cities.

Local authority ‘Old’ 2008 population Revised 2008 population Population increase
Westminster 236,000 246,600 10,600 (4.5%)
Harrow 216,200 225,400 9,200 (4.2%)
Leicester 294,700 303,800 9,100 (3.1%)
Manchester 464,200 473,200 9,000 (1.9%)
Leeds 770,800 779,300 8,400 (1.1%)
Hounslow 222,600 230,200 7,700 (3.5%)
Richmond upon Thames 180,100 187,200 7,100 (3.9%)
Lambeth 274,500 281,400 6,900 (2.5%)
Barnet 331,500 338,100 6,600 (2.0%)
Tower Hamlets 220,500 226,800 6,300 (2.9%)

Table 2 shows the 10 local authorities with the biggest fall in overall population under the revised population estimates. These areas are typically characterised by high levels of migration and/ or large student populations.

Local authority ‘Old’ 2008 population Revised 2008 population Population decrease
Brent 270,600 254,500 -16,100 (-6.0%)
Kensington and Chelsea 180,300 171,100 -9,300 (-5.1%)
Camden 235,700 226,500 -9,200 (-3.9%)
Oxford 153,900 146,500 -7,400 (-4.8%)
Newham 249,500 242,400 -7, 100 (-2.8%)
Colchester 181,000 174,300 -6,800 (-3.7%)
Exeter 123,500 118,500 -5,000 (-4.0%)
Harrogate 160,500 156,100 -4,400 (-2.7%)
Lancaster 143,700 139,500 -4,200 (-2.9%)
Cambridge 122,800 118,700 -4,100 (-3.3%)

The City of London is the local authority most affected in terms of the proportional change, with the total population estimated to be 43% higher on the basis of the revised population estimates. However, the City of London LA has a small population, and the overall increase for the 2008 year is 3,400 (from 7,900 using the old estimates, to 11,300 under the revised estimates).

Population shifts (as a proportion of total population) in other local authorities are significantly smaller, with only 10 local authorities experiencing a 3% or greater increase in population as a result of the revisions, and 10 local authorities experiencing a 3% or greater decrease in population.

What are the geographical patterns?

The map shows the impact of the revised population levels across England, highlighting those areas with the biggest increases (blue) and decreases (yellow).

Impact of revised population levels across England

Impact of revised population levels across England

The largest increases (shaded blue) are seen in metropolitan areas, including a number of inner and outer London boroughs and larger cities. The largest decreases (shaded yellow) are seen in more rural areas, including parts of East Anglia and the rural East Midlands.

Which neighbourhoods are most affected?

The population revisions also affect the estimated populations of small areas such as Super Output Areas, with some neighbourhoods affected very significantly. For example, one LSOA in Howdenshire ward in East Riding of Yorkshire shows an 88% increase in population from the revised estimates; by contrast, an LSOA in Dale (also in East Yorkshire) has 41% fewer people using the revised estimates.

Across England, there are 28 LSOAs with increases of more than 20% as a result of the population revisions (including each of the LSOAs in the City of London), while 14 LSOAs have decreased in population by more than 20%.

For further analysis and briefing of how the revisions affect your area, including impact at the neighbourhood level, please contact us.

The population revisions could affect how initiatives to tackle worklessness and other issues are targeted

Local programmes are often targeted at those areas with the highest levels of worklessness or other priority issue. The mid-year population estimates are a critical component in this process, as the population data is needed to calculate (for example) the proportion of working-age people in receipt of workless benefits. Revisions to the population data therefore can affect the levels of worklessness, with both local authorities and small areas affected.

For example, the revised population in the London Borough of Brent is estimated to be 6% smaller than previously identified under the national estimates. Using the revised population data to calculate the proportion of working age people receiving DWP benefits (a key indicator of deprivation), shows that the estimated Brent value for August 2008 changed from 15.5% to 16.9% using the revised population estimates.

This is a fairly significant change. As a result, Brent’s position ranked against all local authorities in England changed from the 85thst highest (of 326) in terms of proportion of people receiving benefits, to the 62nd highest using the revised population estimates. In total, 30 local authorities across England show changes in rank position of 10 places or more based on the proportion of working age DWP benefit claimants. Small areas are affected to an even greater extent.

Where programmes are targeted at those areas with the highest levels of worklessness (or other indicator), large changes in rank position could affect whether or not areas receive funds. For example, the Neighbourhood Renewal Fund (and later Working Neighbourhoods Fund) were targeted at the most deprived 88 or so local authorities. The Brent example shows how the population revisions can shift local authorities in (or out) of the target group, resulting in large changes to funding levels.

In addition, changes to the population denominators can affect trends over time, with some areas even reversing the direction of travel. Using the Brent example again, the proportion of people receiving DWP benefits fell from 16% to 15.5% over the period 2001 to 2008 using the older population estimates. However, based on the revised population estimates, the proportion of people receiving benefits across Brent actually increased over the same period, from 16% to 16.9%. In other words, the population revisions can show a very different story in terms of the overall direction of travel than older data suggests. In total, 19 local authorities across England saw changes in the direction of travel as a result of the revisions.

Data notes

  • The Office for National Statistics has an ongoing programme of work to improve the mid-year population estimates, the standard measure of population in between the 10-yearly census (next due in 2011). Several improvements to methodology have been made, which affect how the national population is split between local areas (the overall national population remain unchanged).
  • The new methods have resulted in revised estimates for the years 2002 to 2008 for regions, local authorities and small areas. Revised mid-year estimates for local authorities were published in summer 2010, with revised small area estimates published in September.
  • For details of the mid-year population estimates, see http://www.neighbourhood.statistics.gov.uk/dissemination/Info.do?page=news/newsitems/16-september-2010-local-area-population.htm

Stefan Noble, OCSI.


1 Comment » for Impact of revisions to the mid-year population estimates
1 Pings/Trackbacks for "Impact of revisions to the mid-year population estimates"
  1. […] The overall national population has not been changed, only the way that this is split between local authorities and small areas. Elsewhere we’ve published analysis on the changes at local authority level (published on our OCSI blog). […]

Leave a Reply

Your email address will not be published. Required fields are marked *

*