Objectively assessed need and housing targets PAS Spring

Objectively assessed need and housing targets PAS Spring Conference March 2015 Cristina Howick Peter Brett Associates LLP

Introduction • NPPF says • Planning authorities must meet objectively assessed need (‘the OAN’) in full • For all development including housing • Unless they are prevented by supply constraints • Recognised in the Statement • In which case they should export the need to other places • Broadly speaking it’s obvious what they’re after • But in practice it’s challenging • So here’s some advice • • Based on hard experience Highlights from the PAS advice note Including update coming up Work in progress

What is housing need?

We do need (sic) a working definition • Objectively assessed housing need is central to planning • Yet neither NPPF nor PG tell us what it is • ‘Need’ is a very broad term • In PPG 3 (see Annex) it meant something else entirely • There are endless projection scenarios • You can’t tell which is the right one • Unless you know what you’re looking for

A pragmatic definition • ‘The housing that households are able to buy or rent • Either from their own resources • In the market sector • Or with assistance from the State • In the affordable sector or with subsidy’ • ‘Overall housing need’ in the PPG • The market housing that households want and can afford • Plus the affordable and subsidised housing that the State wants and can afford to provide for them • AKA ‘demand’ • Unconstrained • What would be built if planning did not restrict land supply

Caveats • Principle • It’s not an official definition • There isn’t one of those • The PG says ‘Need is… what is likely to be needed’ • It’s imperfect of course • Can’t capture every nuance of the Government’s thinking • Practice • There’s no evidence about unconstrained (policy-off) demand • Planning has been constraining development since 1947 • The best we can do is estimate policy-same (‘policy-neutral’) • What would happen if the constraint stayed the same • That’s what projections try to do • In theory the OAN can be a range • In practice you’re better off with a single number • PG says it’s not an exact science • Official projections are festooned with health warnings • Inspectors sometimes remember this • More often they don’t

Method overview

The logical thread • Start from demography • Then adjust it • Need & target aren’t the same • Separated by the ‘policy line’ • The line is fuzzy • Especially in the PG • Constraints are below • Including policy constraints • Including Green Belt • We now think affordable is below too

The housing market area

NPPG and PG • The Framework says need should be assessed for HMAs • • • What is an HMA? • • • Not proportionate and not reasonable? If each authority draws an area centred on itself • • • Including migration Aim for ‘typically’ 70% self-containment (‘closure’) Sort of suggests each authority should draw their own HMA • • A reasonably self-contained area Both for migration and travel to work PG points to many kinds of data • • • ‘Where market areas cross local authority boundaries’ Which means almost everywhere It will just pull in its nearest neighbours You’d get as many HMAs as authorities We need a top-down method • • That maximises self-containment across the country And centres HMAs on large towns / cities

Drawing the boundary • Start from the CURDS research for NHPAU / CLG • A rigorous and nationally consistent set of HMAs http: //www. ncl. ac. uk/curds/research/defining/NHPAU. htm • Or pre-existing partnerships / relationships • Such as Leps • But don’t exclude anyone just because they’re not old friends • Check the above against the latest data • Calculate migration containment • Commuting also important • If in doubt go for the larger area • Excluding places may get you into trouble • Also consider ‘related areas’ round the edge • (Everything has edges) • They’re your most likely trading partners (import or export) • But will also have other partners • Anyway HMAs often overlap and Inspectors know it

An example ‘Greater Birmingham’ HMA

No perfect answer • Be pragmatic • • Don’t cut across local authority areas • • Makes sense and is massively easier Look beyond the HMA too • • • Too complicated and the answer is still imperfect Ideally HMAs and ‘functional economic areas’ should coincide • • There’s a hierarchy of HMAs • Degrees of self-containment But we have to draw a line somewhere Go for what’s manageable and looks reasonable The Duty to Co-operate doesn’t stop at the HMA boundary You may be asked to import need from conurbations far away Take London, Brighton & Hove, Birmingham • Looking to export over long distances • Maybe indirectly as demand ripples out Whatever you do it’s horribly messy And there isn’t a single answer

Demographic projections

The official demographic projections • Two publications • ONS sub-national population projections (SNPP) • CLG household projections - turn that population into households • Net new households measure housing demand / need • After small adjustment for vacant and second homes • Usually c 3% but more in holiday areas • See ONS Table KS 401 EW • Trend-driven • Roll forward past demographic trends from different ‘base periods’ • Natural change (births less deaths) • Migration • Household formation (HRRs, headship rates) • By demographic group (age x sex x relationship status) • For migration the base period is only five years • Makes the official projections very unstable • And recent projections lock in the recession • Assume no change in non-demographic factors • Including impact of economic climate on housing demand

Recent releases • 2008 -based – ONS 2008, CLG 2008 • Based on very old information • Long before 2011 Census • Interim 2011 -based – ONS 2011, CLG 2011 • Now superseded by • 2012 -based – ONS, CLG • • CLG household projections published 26 Feb 2015 Methodology published 4 th March We’re still trying to work it out Will put an end to our endless number-crunching? • Don’t count on it

First time for at least 200 years that household size did not fall between Censuses

Unattributable population change • People who were in given places in 2011 but not in 2001 • (or vice-versa) • But aren’t accounted for by estimated births / deaths / migration • The national total is 103, 700 • Balance of larger positives and negatives for different places • It’s probably unrecorded or misrecorded migration • To everyone’s surprise ONS / CLG 2012 ignores it • Can make a big difference to demographic projections • Often in large urban areas • What to do about it? • I now think leave it out • Unless you have good local evidence to the contrary

Building demographic scenarios • Start from official projections • Consider alternative scenarios to • Incorporate latest data • The official projections often lag behind • Fix any anomalies in the modelling • Sense check projected future against the past • Look at UPC • Test alternative base periods for migration • Would 10 years be more typical of the long-term trend? • Did unusual things happen in your base period? • Model household formation recovering from the recession? • Until last week ‘indexed’ or ‘blended’ was popular with Inspectors • New 2012 -based CLG maybe a much better answer

What not to do • Derive housing need from nil-migration scenarios • The NPPF says the OAN includes migration • In some places most population growth is net in-migration • Demographic projections roll it forward • And in the market sector we can’t stop it • Richer migrants will displace poorer locals • Or from supply-constrained (‘dwelling-led’) scenarios • Supply constraints are nothing to do with need • They bear on the target but not the need • NPPF, PG and Inspectors are at one on this • Or from old RSS targets • Supply-constrained and policy-led • Or from previous housing completions • The NPPF wants us to build more houses than before

Past supply and market signals

Reading the signals ‘PG: [Household] formation rates may have been suppressed historically by undersupply. . . The assessment will therefore need to reflect the consequences of past under-delivery. As housing projections do not reflect [past] unmet need, LPAs should take a view based on the available evidence of the extent to which household formation rates are of have been constrained by supply. ’ • What evidence? • Past delivery • Related to past planning • Was development restricted by lack of planned supply? • Market signals • Mainly prices/rents higher or rising faster than normal • So how do we OA the N ? • Apply uplift to the projection • PG doesn’t say by how much • Inspectors have made a rule of thumb • 10% for small problem • Suggests 15 or 20% for bigger problem?

Aligning jobs and housing

What not to do • Buy a job forecast • Translate the jobs into population into households & houses • Assuming fixed commuting ratios • Infer number of dwellings ‘needed’ • Why not? • The forecast already assumes a given population • And implicitly households and dwellings • But not the same as you calculate at the end • Usually fewer • Because the forecaster’s jobs-to-population factors are different • Especially commuting ratios are not fixed • In real life commuting adjusts to supply-demand shifts • Objectors have to ask just one question • Show us the population figures behind your job forecast please • And your numbers go in the bin

A real-life example • Experian forecasts 31, 000 net new jobs in the plan period • They also shows population growth • In small print at the bottom • 45, 000 extra residents • Taken from ONS 2010 -based projections • • • The planners didn’t look at that They calculate that 31, 000 new jobs ‘needs’ 74, 000 new residents Hence (say) 32, 000 new houses Makes no sense Happens all the time

What to do • Look at job forecasts • Assuming the correct population • As per preferred demographic projection • Audit the forecasts • Do you believe them? • Are they consistent with policy aspirations? • Are the commuting implications • Credible? • Acceptable / policy-friendly? • If Yes • End of (almost) • Make sure employment land policies use the same numbers

What to do continued • If No • Produce new job numbers (demand) • ‘Realistic as well as aspirational’ • Don’t get carried away • But falling jobs / workforce isn’t on either • Translate into resident population > housing • • Not necessarily fixed commuting ratios Work with the forecaster for an intelligent view Understand how their jobs-to-population relationship works The models vary • If you don’t like the result • Reconsider the job number • In short • Plan for enough housing • To support the jobs you are expecting and planning for

Affordable housing

A different kind of number • • The PG (paras 022 -029 ) shows how to assess affordable need A separate and different calculation to ‘overall need’ (the OAN) Affordable need is not a part of the OAN It’s a different meaning of ‘need’ • About aspiration / entitlement • • • Moves everyone in ‘unsuitable’ homes to suitable ones Not limited to what affordable providers can pay for About ‘ought’ Unlike the OAN and projections They’re about ‘is’ • Not all of it is net new homes • Includes a backlog of existing households • When they leave their unsuitable (for them) homes • Other people can move in • So affordable need is usually a large number • Sometimes larger than the OAN

What to do i. ii. Calculate the OAN as above Estimate how much of it could be new affordable housing • Given expected developer’s contributions % iii. Assess affordable housing need • As per paras 022 -029 of the PG iv. Compare (iii) with (ii) v. Are you meeting enough of the affordable need? vi. If not, consider a higher number • Changes the target, not the OAN

A bad argument • ‘Affordable need as per paras 22 -29 is part of the OAN’ • It can’t be! • It’s a different meaning of need (see above) • Technically we don’t know how to deal with it • The demographic projections already include an affordable element • Though we can’t measure it • So if we add the two we’d be double-counting • But the High Court says otherwise? • See Satnam Millenium Ltd v Warrington Borough Council 19 Feb 2015 • Dangerous

Another bad argument • OAN equals affordable need (paras 22 -29) times S 106 affordable %’ • Say affordable need is 1, 000 dpa and S 106 % is 33% • Your OAN is 3, 000 dpa • Even if demographic projections + adjustments say 1, 200 dpa • This can’t be right • This is another meaning of ‘need’ again • You’d be hugely oversupplying need / demand for market housing • Especially in places where the market is weakest • (low S 106 %) • It wouldn’t get delivered • Smart Inspectors know all this • Let’s hope judges get it too

From housing need to plan target

What to do • Assess supply capacity / constraints • Needs armour-plated evidence • Consider cross-boundary unmet need • Can and should you import? • Can you justify exporting some? • Consider sustainability and well-being • • • Will your proposed target deliver sustainable settlements? With a healthy age structure? Enough critical mass to support key services? Etc etc Remember the OAN is a minimum • It’s fine to plan for more housing • Don’t add the ‘shortfall’ inherited from the last plan • New projections and new plan restart the clock • See Zurich Assurance v Winchester

Final thoughts

Practical • Keep it simple • There’s no need for 18 scenarios and 250 pages • Too many numbers lose sight of the main plot • It’s about a positive strategy for your area • While doing your bit for the larger community

Blue sky • What would a better NPPF / PG look like? • Sort the larger-than local thing • Think again about affordable need • Is the answer more and more market housing? • Much clearer technical guidance • ‘Need’ v ‘demand’ • What to do about projections • Etc etc • It’s not rocket science • Can we learn from European partners? • Is France doing something right for once? • In 2014 its housebuilding fell below 300, 000 • For the first time since records began • England averages 146, 000

Thank you chowick@peterbrett. com