Comment

Where next for the substandard method of assessing housing need?

The Competition and Markets Authority (CMA) published its report on the housebuilding market last month, emphasising how several aspects of our planning system have contributed to the long-term undersupply of new housing relative to need. With the approach to setting housing targets cited as one area with “sub-optimal outcomes”, Andrew Lowe and Antony Pollard reflect on the current situation and explain how housing stock can form a reliable foundation for a new approach.

The CMA rightly notes – on page 150 of its report – that the standard method of assessing housing need in England has been ineffective and “subject to criticism”, due to its continued use of 2014-based household projections and its incorporation of an urban uplift that has been largely resisted by the affected areas.

The Government acknowledged these views in its recently published response to consultation on reforms to national planning policy, but it has committed to reviewing the approach to assessing housing need only when new household projections are released. This will not be until some point in 2025. Do we really need to wait that long? Is there an alternative approach that would boost housing delivery and can be implemented sooner?

Household projections: missing the mark

In its report, the CMA suggests that “a more effective methodology” for creating targets that accurately reflect need would be:

• Easy to understand;
• Based on reliable evidence;
• Regularly applied but stable;
• Largely free of adjustments linked to broader policy aims; and
• Aligned with a national target.

There can be no guarantee that a new approach based on new household projections would satisfy these criteria. While ostensibly based on recent trends – easy to explain on that front – there is an inherent complexity to demographic projections, due to the myriad assumptions made in their development on everything from levels of international migration to life expectancy and household formation rates. The Office for National Statistics (ONS) – which produces the official projections – increasingly acknowledges the sensitivity of these projections to various assumptions, by altering them in published variants and emphasising that its projections are in no way “predictions or forecasts”, nor “a measure of how many houses would need to be built to meet housing demand”.

There is equally always the spectre of new projections on the horizon, where the ONS typically updates them every two years. The results have recently been unpredictable, with the two releases since the 2014-based projections having first generally reduced growth in individual areas – by an average of 22% – before increasing it again by a coincidentally identical average of 22%, albeit without returning to original levels. Such volatile projections create an extremely unstable foundation for any method of assessing housing need, as the current Government implicitly accepts in continuing to favour the “stability, consistency and certainty” offered by 2014-based household projections.

Over five years have passed since this quick fix was first made, and it is fair to say that this hastily-applied sticking plaster is now coming loose. The ONS has produced population estimates for no less than eight extra years, since 2014, revealing that few areas have close to the number of residents assumed by the projections that underpin the standard method. Out of 309 local authorities, only 58 – less than a fifth – had a population, in 2022, that was within 1% of that assumed by the 2014-based projections. They were at least 5% out in some 65 areas.

This all suggests that the next round of projections will once again be quite different to those that currently feature in the standard method. With new – and as yet unspecified – assumptions on household formation to also be thrown into the mix, their impact at the local level is anyone’s guess at this point. It would also take a giant leap of faith to assume that they will line up with the enduring national target for 300,000 homes every year, without upward adjustments of the kind that are often perceived to move away from the “underlying housing need” according to the CMA report.

An alternative approach: taking stock

All of the above adds doubt as to whether the forthcoming household projections can indeed play a prominent role in a revised standard method, which addresses existing deficiencies.

Having worked with our clients over recent years to explore numerous alternatives – incorporating everything from, yes, household projections to past delivery and even job density – we believe that the existing, occupied housing stock of an area provides a more reliable foundation for a new approach, that would specify the minimum rate at which stock should grow each year to deliver at least 300,000 homes nationally.

Integrating stock would lead to a method that better adheres to the principles set out by the CMA. By relating more closely to the existing population of an area a stock-based approach would avoid perception of “top-down” targets and more quickly garner community and political understanding. While no standard method is likely to gain universal support, it is notable that the Home Builders Federation – representing the country’s largest housebuilders – has recently advocated the benefits of a stock-based starting point over projections.

The stock-based approach is very simple, fairly assuming that every area makes an equal contribution towards meeting housing need by, for example, providing one home each year for every 100 that are currently occupied^[1]. It could be based on credible and publicly available Council Tax data, which is updated annually and consistently recorded bringing little risk of volatility. It can also be designed in such a way as to minimise the need for centrally-imposed adjustments, akin to the urban uplift, while recognising that this would only provide a starting point with plan-makers subsequently applying adjustments that reflect local circumstances. These would respond to local affordability issues, specific demographic challenges and economic investment, for example.

Crucially, such an approach can be fully aligned with a national target via the selection of a minimum growth rate, with a figure of 1.2% currently generating outcomes that would total almost 300,000 homes per annum. This would provide a firm foundation from which local adjustments can be applied.

We believe that an approach of this nature could be implemented very quickly through updated policy and guidance, without needing to wait for new household projections that could yet be delayed and, even when available, could fail to provide a reliable basis for assessing housing need throughout England. A new approach, incorporating housing stock, would contrastingly introduce much-needed certainty and allow local authorities to bring forward Local Plans that fairly contribute towards the national drive to deliver 300,000 homes per annum.

In the meantime...

For all its flaws, we clearly must work with the current standard method until it is reviewed. Its outcomes are never fixed and changed, for most areas, on 25 March 2024 with the release of new affordability ratios. As in previous years, we have collaborated with the Land, Planning and Development Federation (LPDF) to determine how this new data affects the minimum need suggested for every local authority in England. Download our new report below.

To continue the conversation or for further details and expert advice, reach out to Andrew Lowe or Antony Pollard.

14 March 2024

[1] If the housing stock is assumed, in this example, to grow by a minimum of 1.0% per annum