The impact of immigration on housing in England

26 March, 2007

Summary
1. New government household projections show an annual rate of household growth of 223,000 a year between 2004 and 2026. 73,000 (a third) of this increase is due to net migration into England. This amounts to a requirement for 200 new homes every day to house the additional immigrant population.

2. The household projections are based on population projections which assume long-term net migration rates of 130,000 into England. The Treasury use higher net migration rates in determining their long-term trend economic growth rate. If migration were to follow the Treasury’s migration assumptions the household growth attributable to migration would be about 92,000 households a year or 250 per day.

3. The annual rate of increase of housing stock in England in the period 2000 to 2005 has been close to household formation rates projected in 2000. However, net migration rates in this period have been about 85,000 a year more than the rates used in those projections. This will have resulted in household formation rates exceeding additions to housing stock by 40-45,000 a year. This will have been a major driver of the sharp increase in house prices.

4. If net migration into England were brought into balance the need for additional house building identified in the Barker report would be largely removed. Futhermore, much of the building on greenfield sites would be rendered unnecessary because a much greater proportion of new housing could be built on brownfield sites.

The new household projections

5. Updated household projections for England were released by the Department for Communities and Local Government (DCLG) on 16th March 2007 [1].

6. The projections show an annual rate of household [2] growth in England from 2004 to 2026 of 223,000 – an increase of 14,000 on the previous projections. Of the 223,000 households, 33% or 73,000 per annum are attributed by the DCLG to the assumed level of net migration into England. The other 150,000 a year projected growth in households results from growth in the adult population (other than from immigration), changes in the age distribution of the population (i.e. more pensioners forming more single person households) and social changes resulting in more people living alone. The government can only directly influence one of these factors – namely immigration. In the 22 years from 2004 to 2026 a total of 1.6 million households are projected to form as a result of migration. To put this into perspective, at the 2001 census the city of Birmingham had a total of 404,000 household spaces, so 1.6 million households equates to 4 cities the size of Birmingham in just 22 years.

7. These DCLG projections use the latest population projections produced by the Government Actuary’s Department (GAD) which assume the following levels of net migration into England:



8. The long-term assumption for England of 130,000 a year is made up of 138,000 net international migration less a net annual outflow of 8,000 people from England to the rest of the United Kingdom.

9. It may be suggested that the impact of immigration on housing demand is mitigated by the fact that some immigrants live in multi-occupation, at least in their early years. To test this, we commissioned a table from the 2001 census showing households distinguished between the foreign born and UK born. The number of persons per household was virtually the same. [3] Any such effect is, therefore temporary and, since large scale immigration has been taking place for some years, the impact on housing demand will already be coming through.

Treasury assumptions on international migration

10. As can be seen the projections assume that international migration to the UK will slow rapidly - indeed that it will fall by 40% from its peak level over three years and stay at that lower level for the following 18 years. This assumption was challenged in December 2006 by the Treasury in a pre-budget report [4] entitled ‘Trend growth, new evidence and prospects’ in which they argued that trend growth after 2006 would be 2.75% rather than 2.5%. Part of the rationale for the increase was that the Treasury anticipated net migration to the UK as a whole falling less than GAD’s assumptions to a long-term level of around 185,000 a year instead of 145,000 [5].

11. Assuming England maintains its 90% share of overall net migration, 185,000 net migration into the UK as a whole would equate to about 166,000 a year net migration into England.

12. The DCLG projections include a ‘high migration’ variant of 190,000 a year net migration into England which would result in 105,000 households being formed through migration each year. Net migration of 166,000 a year will therefore result in about 92,000 households being formed each year through migration or about 250 houses a day. This amounts to over 2 million households[6] in the 22 year projection period – that is five cities the size of Birmingham.

13. Such projections are, necessarily, based on assumptions about future events but it is important that government departments make consistent assumptions. Otherwise there is a disconnection between, in this case, Treasury plans about revenue growth and DCLG expenditure on house building and associated infrastructure.

Previous household projections

14. The two previous sets of household projections were produced in 2000 (using 1996 as a base year) and 2006 (using 2003 as a base year). The 1996-based projections assumed net migration of 65,000 a year. At this rate migration would have contributed about 28,000 households annually out of a total projection of 152,000 a year.

15. The 2003-based projections assumed net migration into England of 124,000 a year. Migration accounted for 65,000 of the 209,000 households projected to be formed each year.

16. In fact in the calendar years 2000 to 2004 inclusive net international migration to England averaged over 170,000 a year [7]. 2005 data is not yet available but it is likely that the total will be at a similar level. This has been partly balanced by a net outflow of people from England to the rest of the UK averaging 18,000 a year[8] giving overall net migration into England of around 150,000 a year. Migration levels have therefore exceeded the 1996-based projections by about 85,000 a year – resulting in the addition of 40-45,000 households a year.

17. The actual increase in the housing stock in the period 2000 to 2005 inclusive was 877,000 units – an average of 146,000 a year – just a little less than the 1996-based household projections. Put another way, housing plans based on the 2000 projections would have been almost sufficient to meet demand had it not been for the massive increase in immigration in subsequent years. We conclude therefore that migration levels since 2000 have been a very significant factor in demand for housing exceeding supply. This in turn will have fuelled the very large rises in house prices seen in recent years.

18. The increase now being achieved in the housing stock would be virtually sufficient to satisfy the demand for housing (150,000 a year) if net migration into Britain were brought into balance. This would largely remove the need identified in the Barker report for a 50,000 a year increase in the house building programme (apart from the need to satisfy pent up demand resulting mainly from high levels of migration over the last few years).

19. In 2005, 74% (about 120,000) of new dwellings were built on brownfield sites out of a total of 163,000 new housing completions. If it proves possible to maintain the numbers of dwellings being built on brownfield sites at 120,000 the new household projections would indicate that about 100,000 dwellings each year would require greenfield sites. In contrast, with zero net migration, the number of greenfield site dwellings needed would be about a third of this level at just 30,000 a year.

---