News Briefing - Misreading the tea leaves
10 March 2009
Making sense of the household projections
1. The Government is expected to publish new projections tomorrow (Wednesday 11 March) for the number of households expected to form in England over the next 20 years. These projections will inform the national debate about how many new homes we need and where they should be built. CPRE has a keen interest in how projections are put together and interpreted because new homes — especially those built on greenfield land — have a huge impact on the countryside and the environment.
2. The previous set of projections, published last year, were based on 2004 population projections. They suggested the number of households in England would grow from 21 million in 2004 to 26 million in 2026 — on average 223,000 additional households were expected to form each year.
Trend-based projections
3. The household projections are just that: they project recent historical trends into the future. They are not a reliable indicator of what will actually happen in the future or of housing need. As John Prescott MP, then deputy Prime Minister, remarked ten years ago:
‘The projections are not forecasts, estimates or predictions. They are based entirely on what might be expected to occur if previous trends continue and are heavily dependent on the assumptions involved. Such trends can and do change as a result, for example, of demographic or economic factors.’
(Hansard, 29 March 1999, column 471)
4. CPRE is concerned about the dangers of relying solely on projections to determine the level and location of new housing development. Projected growth in the number of households does not tell us how many homes need to be built. What matters is the relationship between the number of households and the total housing stock. The Government is aware of the limitations of the projections, as the following statement illustrates:
‘The household projections are not an assessment of housing need. They do not take account of future policies. They are an indication of the likely increase in households given the continuation of recent demographic trends. They are one part of the evidence that Regional Planning Bodies and local authorities use in the assessment of future housing requirements.’
Communities and Local Government, Statistical Release: Revised projections of households for the English regions to 2026, 28 February 2008, note 4 to editors.
How the projections are calculated
5. Projections show what would occur if recent trends in rates of birth, death, migration, marriage, co-habitation and divorce were to continue into the future. But the assumptions underlying these projections are highly questionable and may not hold true. This can significantly affect the resulting figures: in the late 1990s the number of people co-habitating proved larger than had been assumed. A wide range of factors affect household formation but do not form part of the method used for calculating projections. These include interest rates, household incomes, welfare benefits, national and local policy, lifestyles and the cost and availability of housing. Economic conditions have a strong bearing on migration — during a downturn immigration usually falls due to less attractive employment prospects. The turbulence in the wider economy and uncertainties surrounding future prospects, highlight the importance of testing projections in the light of realistic assumptions about future economic performance.
6. Moreover, the supply of housing influences the demand for it. Constructing new housing in an area encourages in-migration, which can lead to further demand in the future. While this benefits some places, it can harm others. In rural areas with major environmental constraints and a lack of affordable housing, local people are being squeezed out of the market where they cannot compete with wealthier households moving in from more prosperous areas. Policy interventions, or a lack of them, perpetuate trends which become self-reinforcing.
7. CPRE does not accept the unsound assumption that we should plan the future on the basis of what has happened in the past. Predicting household growth and setting targets based on this is inherently problematic, as David Miles, Chief Economist at Morgan Stanley explained to the Treasury Select Committee last October:
‘I think that the amount of uncertainty about the scale of household formation is actually extremely large and it is not just uncertainty about what the scale of net migration into the UK might be. It also touches on issues […] about longevity, people’s changing patterns of sharing homes, changing attitudes towards home ownership and the risk, rather than just the upside of owning your own home. And also when people choose to leave the family home. […] all these things are affected by conditions in the housing market and the economy and they are all going to change in ways that I think it would be extremely difficult to forecast right now. I think it is one reason why having some kind of inflexible target for how many homes we need to build is a very dangerous road to go down.’
Plan, monitor and manage
8. CPRE believes that in responding to household growth public policy should become less dominated by trend-based projections and take greater account of the implications for future lifestyles, quality of life and the environment. We believe the way forward lies in adopting a ‘plan, monitor and manage’ approach along the lines called for by John Prescott a decade ago:
‘We want to replace the top-down, “predict and provide” mentality of the past with a more responsive and accountable system that is better able to revitalise our towns and cities and protect a living countryside that we all enjoy.’
(Hansard, 23 February 1998, column 24)
9. A return to a ‘predict and provide’ approach based on mechanical projections of past trends would lock us into a cycle of urban decline and countryside sprawl. Slavishly following projections and setting targets based on them is potentially harmful even at greatly reduced housebuilding levels, as in the present downturn. This is because planning authorities must allocate land to accommodate a national target of three million homes by 2020 (240,000 homes a year by 2016). This target is above the projected rate of household formation and unlikely to be achieved.
10. Developers will naturally respond to this land surplus by ‘cherry picking’ more profitable greenfield sites in preference to brownfield locations. By contrast, a ‘plan, monitor and manage approach’ looks to meet housing needs by making better use of existing property; considers wider objectives, such as the capacity of the environment to accommodate development; monitors what happens on the ground and takes action in the light of this.
1. The Government is expected to publish new projections tomorrow (Wednesday 11 March) for the number of households expected to form in England over the next 20 years. These projections will inform the national debate about how many new homes we need and where they should be built. CPRE has a keen interest in how projections are put together and interpreted because new homes — especially those built on greenfield land — have a huge impact on the countryside and the environment.
2. The previous set of projections, published last year, were based on 2004 population projections. They suggested the number of households in England would grow from 21 million in 2004 to 26 million in 2026 — on average 223,000 additional households were expected to form each year.
Trend-based projections
3. The household projections are just that: they project recent historical trends into the future. They are not a reliable indicator of what will actually happen in the future or of housing need. As John Prescott MP, then deputy Prime Minister, remarked ten years ago:
‘The projections are not forecasts, estimates or predictions. They are based entirely on what might be expected to occur if previous trends continue and are heavily dependent on the assumptions involved. Such trends can and do change as a result, for example, of demographic or economic factors.’
(Hansard, 29 March 1999, column 471)
4. CPRE is concerned about the dangers of relying solely on projections to determine the level and location of new housing development. Projected growth in the number of households does not tell us how many homes need to be built. What matters is the relationship between the number of households and the total housing stock. The Government is aware of the limitations of the projections, as the following statement illustrates:
‘The household projections are not an assessment of housing need. They do not take account of future policies. They are an indication of the likely increase in households given the continuation of recent demographic trends. They are one part of the evidence that Regional Planning Bodies and local authorities use in the assessment of future housing requirements.’
Communities and Local Government, Statistical Release: Revised projections of households for the English regions to 2026, 28 February 2008, note 4 to editors.
How the projections are calculated
5. Projections show what would occur if recent trends in rates of birth, death, migration, marriage, co-habitation and divorce were to continue into the future. But the assumptions underlying these projections are highly questionable and may not hold true. This can significantly affect the resulting figures: in the late 1990s the number of people co-habitating proved larger than had been assumed. A wide range of factors affect household formation but do not form part of the method used for calculating projections. These include interest rates, household incomes, welfare benefits, national and local policy, lifestyles and the cost and availability of housing. Economic conditions have a strong bearing on migration — during a downturn immigration usually falls due to less attractive employment prospects. The turbulence in the wider economy and uncertainties surrounding future prospects, highlight the importance of testing projections in the light of realistic assumptions about future economic performance.
6. Moreover, the supply of housing influences the demand for it. Constructing new housing in an area encourages in-migration, which can lead to further demand in the future. While this benefits some places, it can harm others. In rural areas with major environmental constraints and a lack of affordable housing, local people are being squeezed out of the market where they cannot compete with wealthier households moving in from more prosperous areas. Policy interventions, or a lack of them, perpetuate trends which become self-reinforcing.
7. CPRE does not accept the unsound assumption that we should plan the future on the basis of what has happened in the past. Predicting household growth and setting targets based on this is inherently problematic, as David Miles, Chief Economist at Morgan Stanley explained to the Treasury Select Committee last October:
‘I think that the amount of uncertainty about the scale of household formation is actually extremely large and it is not just uncertainty about what the scale of net migration into the UK might be. It also touches on issues […] about longevity, people’s changing patterns of sharing homes, changing attitudes towards home ownership and the risk, rather than just the upside of owning your own home. And also when people choose to leave the family home. […] all these things are affected by conditions in the housing market and the economy and they are all going to change in ways that I think it would be extremely difficult to forecast right now. I think it is one reason why having some kind of inflexible target for how many homes we need to build is a very dangerous road to go down.’
Plan, monitor and manage
8. CPRE believes that in responding to household growth public policy should become less dominated by trend-based projections and take greater account of the implications for future lifestyles, quality of life and the environment. We believe the way forward lies in adopting a ‘plan, monitor and manage’ approach along the lines called for by John Prescott a decade ago:
‘We want to replace the top-down, “predict and provide” mentality of the past with a more responsive and accountable system that is better able to revitalise our towns and cities and protect a living countryside that we all enjoy.’
(Hansard, 23 February 1998, column 24)
9. A return to a ‘predict and provide’ approach based on mechanical projections of past trends would lock us into a cycle of urban decline and countryside sprawl. Slavishly following projections and setting targets based on them is potentially harmful even at greatly reduced housebuilding levels, as in the present downturn. This is because planning authorities must allocate land to accommodate a national target of three million homes by 2020 (240,000 homes a year by 2016). This target is above the projected rate of household formation and unlikely to be achieved.
10. Developers will naturally respond to this land surplus by ‘cherry picking’ more profitable greenfield sites in preference to brownfield locations. By contrast, a ‘plan, monitor and manage approach’ looks to meet housing needs by making better use of existing property; considers wider objectives, such as the capacity of the environment to accommodate development; monitors what happens on the ground and takes action in the light of this.
