This website uses cookies so that we can provide you with the best possible experience. If you continue to use this site we will assume that you are happy with this. You can find out more about how we use cookies here. If you would like to know more about cookies, or how you can delete them, click here.

The highs and lows of the housing numbers game

19 January 2016 Author: Neil Harris

How many houses are needed in your area over the next 10 to 15 years? It sounds a simple question, doesn't it? Setting out a figure for the number of houses needed over a given time frame is a key part of any plan.

Yet calculating a housing requirement is a complex task that demands skill and expertise. It is also a task where the stakes for a wide range of interests are especially high. Provide too much or too little land for housing and you risk losing control of the strategy underpinning your plan. The consequences of getting it wrong are very real and practical ones too.

So, where do you start with this complex task? The starting point is often household projections data. Household projections data are used for a great many activities. These range from planning of education and health facilities, through to calculation of local funding and land use planning. The household projections - put very crudely - take a set of recent trends related to population change over the past 5 to 10 years and project these forward into the future over a timescale of up to 25 years.

Planners are very aware of the limitations and caveats of using household projections as a basis for calculating housing requirements. The use of past trends for planning future development inevitably has limitations. Household projections only indicate what would happen if trends in the recent past continue. Planners have a particular concern for the longer-term and so are rightly sceptical about using a short period from which to project a trend into the future.

Our new research for the RTPI on the use of household projections data to produce land use plans in Wales highlights some of the challenges planners face when calculating housing requirements. For example, household projection data released in 2014 identifies very different numbers of expected households for local authorities than the preceding set of household projections. (This research follows work done for the RTPI by Cambridge University on household projections in England).


The simple change of base year from which to project the number of expected new households - from 2008 to 2011 - resulted in a reduction of 50% or more in the number of projected additional households for half of local planning authorities in Wales. In other words, the number of projected additional households can change dramatically depending on the trends occurring in the preceding five years. That simple task of identifying the houses required in your area during the next 10 years just became a little bit more complex.

The very different figures for projected additional households - issued just a few years apart - has led to some critical debate in Wales. There has been discussion about whether the latest figures are in some way 'off trend' as a consequence of recession and its impacts on household formation and migration. Some interests expect a return to pre-recession trends, while others argue that the population changes witnessed immediately before recession are also not a good reflection of longer-term trends.

The problem here comes back to the nature of the household projections themselves. The answer to this problem may lie in developing household projections for land use planning purposes over longer time frames. In other words, planning for the longer-term should be based on data that 'smooths out' the highs and lows of population and household change, and also recognises the distinctive character of land use planning as a longer-term activity. Effective planning is based on looking further back into the past as basis for trend projection, but also engaging in the interesting work of considering alternative futures that may unfold.

The recent patterns of international migration occurring across Europe, for example, present a fresh challenge for planners in the United Kingdom identifying housing requirements in their plans. It is clear that projections of the past only get us so far in planning for the future of our communities.

So, what might planners do to address the challenges of using household projections data as a starting point for assessing housing requirements?

They need to get their hands dirty with data, develop variant projections, work with different assumptions about what might be emerging trends as well as documenting past trends. They then also need to use this data as a basis for steering their communities through the political choices about priorities, as well  as the difficult decisions to be made about where housing land is to be provided if a community's housing needs are to be met.

The skills and expertise needed to do some of this work are in short supply, particularly in relation to demographic expertise and statistical skills, and so there is a need to address this skills gap. Planners may then - even if always cautiously and contingently - be able to identify with greater confidence the number of houses required in their area in the next 10 to 15 years.

Neil Harris is a Chartered Town Planner and a Senior Lecturer in the School of Planning and Geography at Cardiff University. Neil is part of a team of colleagues that recently completed research for RTPI Cymru on the use of household projections in preparing Local Development Plans in Wales. The full report is available here, along with a research briefing.

Neil Harris

Neil Harris

Senior Lecturer, School of Planning and Geography, Cardiff University