Blog: 29 June 2020New Household Projections, same old story?

Dan Usher

Posted by:

Dan Usher

Development Economics Associate

Ebbsfleet office

View blog posts | view profile


Introduction

Today (29 June 2020) the Office for National Statistics (ONS) released their latest official household projections for England. As the ONS state, these projections provide an “indication of the future number of households in England and its regions and local authorities.” The projections are made from a 2018 base year and cover a 25-year period up to 2043. They replace the 2016-based household projections also produced by the ONS. 

Headlines

Overall, the 2018-based series project 16.2% growth in England over 25 years (2018-2043).

The previous 2016-based household projections projected 17.3% growth in England over 25 years (2016-2041). The 2014-based projections used for the NPPF's 'Standard Method' projected 23.1% growth over 25 years (2014-2039).

However, when compared to previous household projections there are regional variations as the table below shows for the current 10-year Standard Method calculation period (2020-2030). It is important to note that these projections are not currently being used for the calculation of the Standard Method as we explain below.

Table 1: Household Projections Comparison

Source: ONS/MHCLG

Use in Planning

From a planning perspective, at the present time these 2018-based projections have little impact on emerging Local Plans and Development Management decisions. This is because Planning Practice Guidance (PPG) underpinning the 2019 National Planning Policy Framework (NPPF) clearly states that the 2014-based household projections (produced by MHCLG and not ONS) should be used for the purposes of the Government’s ‘Standard Method’ calculation of ‘minimum’ housing need.

The PPG was drafted in this way following the publication of the previous 2016-based household projections, the first time that household projections were produced by the ONS, having historically been prepared by MHCLG up to and including the 2014-based series. The 2016-based projections indicated lower household growth than the 2014 series, and negatively affected the Government’s target of building 300,000 homes per annum nationally by the mid-2020s. There were also concerns over the methodology of the 2016-based series.

The 2014 projections were therefore reintroduced for the purposes of calculating the Government’s ‘Standard Method’, and the final PPG (February 2019) made it clear that the 2016 projections could not be used by local planning authorities for departing from the Standard Method. 

Given the 2018-based household projections are lower nationally than the 2016-based series, the PPG may be revised to include guidance that these latest projections are also inadequate for the purposes of Government’s 300,000 homes per annum target. In this context a revised Standard Method and accompanying PPG were expected to be published before these 2018 projections were published, however owing to the Covid-19 crisis this appears to have been delayed.

However, were the 2018-based household projections to be used for calculating the Standard Method using the existing methodology, Table 2 summarises how they would affect the country and the regions.

Table 2 shows how use of the 2018-based household projections would result in a significant decline nationally in respect of the Standard Method calculation. Nearly 60,000 less homes would be required than when calculated against the 2014-based household projections currently used for Standard Method, and this would result in the Standard Method falling nearly 100,000 homes per annum short of the 300,000 homes per annum targeted by Government by the mid-2020s.

Table 2: Existing Standard Method Calculation (2014 and 2018 household projections)

Source: ONS/MHCLG

However, the regions show a contrasting picture. The 2018 projections would result in an increase in capped Standard Method calculations for the East Midlands, North East, North West, and West Midlands. Uncapped, there would be a significant increase in the East Midlands, and marginal increases in the North East, North West, and South West. However these increases would be outweighed by the significant declines in London, the South East, and the East of England.

Methodology

Notwithstanding the minimal impact of these latest 2018 projections on planning, here in the Development Economics team we have our own concerns with the methodology underpinning them. The ‘principal projection’ of the 2018 household projections is underpinned by population growth projected by ONS’ 2018-based Sub National Population Projections (SNPP). It is notable then that the 2018-based SNPP are underpinned themselves by only two years net migration trends. Is this a robust basis for predicting housing need?

Using a two-year migration trend to project 25 years population growth appears inadequate and subject to wild fluctuations. For example, a large development could result in two years of significant in-migration to an area, or out-migration from another, skewing the trend in either direction. 

To put this in context, previous ONS SNPP were underpinned by a five-year trend. Furthermore under the previous 2012 NPPF’s ‘Objective Assessment of Housing Need’ (OAN), many consultants and local authorities argued that a five-year trend wasn’t representative enough and the use of 10 or sometimes 15-year trends should be preferred to ensure that fluctuations in migration did not over or under-estimate population and household growth. We therefore consider that underpinning the new projections with only two years of net-migration is a serious weakness and risks them being inaccurate. The data released by ONS today suggests that ONS themselves do not have confidence in the principal projection, with a 10-year net migration ‘variant’ amongst other variants also published.

The second factor is Household Formation Rates (HFRs). This is the rate at which different age groups in the projections are likely to form their own independent household. Consistent with their previous 2016-based household projections, the ONS have used a much shorter trend period (from 2001 onwards) than previous MHCLG household projections (1971 onwards) to inform their projected HFRs. 

The new ONS approach was questioned by Government in their consultation paper which ultimately recommended use of the 2014-based MHCLG household projections ahead of the 2016 series for the Standard Method included in the 2019 NPPF. We would also question the logic in the ONS approach because it is notable how household formation has been steadily declining in younger age groups since 2001 due to the rapidly worsening affordability situation across the country.

The result of this affordability trend is that more and more young people each year are unable to form their own household and have become ‘concealed’ households/families, living with friends or family. ONS data published in 2014 showed a 71% increase in concealed households nationally between the 2001 Census and 2011 Census. Given the worsening of affordability in many areas since the 2011 Census, this situation is highly likely to have become even more acute in the decade since the 2011 Census took place.

The question to ask is therefore ‘What level of household formation is reasonable?’ in respect of household projections, particularly if these projections have a direct effect on housebuilding. Any approach which exacerbates the problem of concealed households/families in younger age groups would not be fit for purpose.

Summary

The 2018-based household projections published today do not affect the calculation of the 2019 NPPF’s ‘Standard Method’ until any change in the NPPF and PPG is published. Notwithstanding this and given their overall projection nationally, they would only serve to negatively impact the Government’s aspiration to supply 300,000 new homes every year by the mid-2020s under the existing Standard Method.

We also question their accuracy as the basis for predicting household growth, given the very short two-year net-migration trend used to underpin them. Furthermore, the household formation rates assumed in the projections are drawn from a period in which affordability has worsened significantly and younger people have found it increasingly difficult to form their own independent households. We do not consider this to be a sound basis for projecting household growth into the future in the context of Government’s aspiration to build 300,000 homes per annum by the mid-2020s.

 

Posted with the following keywords:
Housing need, planning reform, Household projections, #letsfindaway, build back better