• Issue 41

  • Oct 2018

The Source


There are Lies, Damned Lies and Household Projections

The intentions of Government when it decided to change the way housing needs were calculated in the English planning system were laudable. The Housing White Paper spoke of creating a simpler system, where local authority officers and residents alike could easily understand the way housing numbers were generated in their area. It would stream-line Local Plan preparation, saving time and money for hard pressed Councils. Where we are today feels very different to this panacea.

The simpler system described above took the form of a university calculation (‘the standard method’) for assessing housing needs. The Government published the requirement to use the standard method in July 2018 as part of the revised National Planning Policy Framework (NPPF). This set out how and in what circumstances the standard method should be used for plan-making and decision-taking. The more perceptive of us also noted the Government’s health warning, published alongside the NPPF:

The government is aware that lower than previously forecast population projections have an impact on the outputs associated with the method. Specifically it is noted that the revised [Household Projections] projections are likely to result in the minimum need numbers generated by the method being subject to a significant reduction, once the relevant household projection figures are released in September 2018. In the housing white paper the government was clear that reforms set out (which included the introduction of a standard method for assessing housing need) should lead to more homes being built. In order to ensure that the outputs associated with the method are consistent with this, we will consider adjusting the method after the household projections are released in September 2018. We will consult on the specific details of any change at that time. It should be noted that the intention is to consider adjusting the method to ensure that the starting point in the plan-making process is consistent in aggregate with the proposals in Planning for the right homes in the right places consultation and continues to be consistent with ensuring that 300,000 homes are built per year by the mid-2020s.”

Essentially the moment that the Government committed to the standard method it had committed to a review of it. Those of us familiar with the modern day working of Local Plans will recognise the strategy that the Government has deployed here: adopt something which does not quite do the job and commit to an early review to remedy the deficiencies. Most observers have recognised there is something slightly perverse about the Government’s position. It is essentially saying that the answer is 300,000 homes per annum and they will change the standard method (presumably every time the statistics change) so it meets that number. A rather unusual way to create public policy but as we will see, the Government’s statement on the projections was worryingly prescient.

On the 20th September the Office for National Statistics uploaded a series of spreadsheets to its website. As innocuous as this sounds, the spreadsheets had just wiped off around 50k homes per year from the annual housing needs of England. Hordes of planning consultants publishing charts, infographics and blogs declared 20/09/18 the End of Days and lamented the “Madness in the Method”. The more measured pointed to the Government’s commitment to review the process saying it was too early to comment on the implications of it all just yet.

The change in the figures had come from a significant drop in population projections and a change to the way in which household formation is understood. A big reason for the drop in population projections was an obscure change ONS had made to life expectancy in England. The more controversial change in the figures relates to household formation. Previously Household Projections took a long-term trend (from 1971) when projecting forward household formation rates. The 2016 Household Projections use a different methodology which essentially projects forward the household formation rates which occurred between 2001 and 2011 – a period in which England didn’t build enough homes and a decade when we saw a house price boom, a credit crunch and a global recession. Not many of us hope to see these conditions perpetuated but that is essentially what the latest household projections assume. Even the ONS recognise this issue, demonstrated by the fact they intend to publish:

“… a variant 2016-based household projections in which household formation rates for younger adults (those aged 25 to 44 years) are higher. The purpose of this variant would be to illustrate the uncertainty in the projections around the future household formation patterns of this age group“.

On 26th October, MHCLG published a proposal for consultation which would mean, in the short-term, scrapping the 2016 Household Projections for the purposes of the standard method and reverting to the 2014 Household Projections. In the long-term, the consultation document proposes a review of the standard method formula so it provides stability and certainty for Councils and supports the Government’s objective of delivering 300,000 homes p.a. by the mid-2020s.

So, where are we after all this? The use of a standard method is now enshrined in policy and the 24th January 2019 (when the standard method starts to apply to Local Plan preparation) is fast approaching. The PPG, which was updated in September [See link to a discussion of this], contains a standard methodology but the Government has already proposed a change to it. The Government’s reviewed position (reverting to the 2014 projections) offers a short-term fix but provides no clarity about how the methodology will change to meet long term needs. ONS has also committed to re-running the 2016 household projections due to “uncertainties in […] household formation patterns”. It’s fair to say that the current circumstances don’t compare particularly well when set against the original objectives of the Housing White Paper.

Where do we go from here?

MHCLG have proposed a change to the standard method which involves rejecting the latest data and rethinking the long-term methodology. The consultation runs to 7 December 2018 but the short-term fix will likely provide clarity for planning authorities submitting their plans next year.

In the long-term, the Government has proposed to “establish a new method […] by the time the next projections are issued”. Household Projection are published every two years so there could be some wait before we know what is happening here. If a change is made then a period of consultation will likely follow. We do not know how the standard method will be updated in the long-tern but if it is still to be based on the latest Household Projections then it is likely that some change will be made to the affordability adjustment – making it applicable at lower affordability ratios or increasing the adjustment. We do not anticipate that the 40% cap (that limits the scale of the overall increase) will be removed but there could be adjustments to it to boost numbers. In summary there will continue to be a period of uncertainty.

Savills’ view is that this whole episode, whilst aiming to achieve the right outcomes, has brought starkly into the spotlight the pitfalls of relying too heavily on household projections to assess housing need. The standard methodology has always had three flaws: the overall number is too low (even with the 2014 Household Projections it produces an annual need for around 269,000 dwellings per annum); the need is not sufficiently focused on the most unaffordable regions; and there are some very unaffordable areas with very low housing need due to a track-record of suppressed house building and household formation. Savills’ recommended approach, first published in October 2017 and set out in detail here [link], is more resilient to changes in the household projection methodology. It reduces the impact of suppressed household formation and results in a housing need for England of 290,000 homes per annum, within 4% of the Government’s housebuilding target. The distribution of housing need also responds to the challenge to improve housing affordability. Ultimately it gets us far closer to the future that the Government originally described in the Housing White Paper.


Tom Baker