Guidance

Housing Delivery Test: 2018 measurement technical note

Published 19 February 2019

Applies to England

Introduction

This document sets out the technical process followed in order to calculate the 2018 Housing Delivery Test measurement in line with the published Housing Delivery Test rulebook.

This document should be read alongside the Housing Delivery Test rulebook, and aims to offer more transparency into the process as opposed to setting out the policy in detail.

All terminology in this document mirrors that in the Housing Delivery Test rulebook.

The Housing Delivery Test measurement will be published annually by the department. The Housing Delivery Test period covers the previous 3 financial years; in the case of the 2018 measurement the years are 2015/16, 2016/17 and 2017/18.

In high level terms the Housing Delivery Test compares the net homes delivered over 3 years to the homes that should have built over the same period (their housing requirement).

Housing Delivery Test (%) =
Total net homes delivered over 3 year period
divided by
Total number of homes required over 3 year period

The calculation is carried out based on published information and data specifically collected by the department from local planning authorities, National Parks and Development Corporations for the purpose of calculating the Housing Delivery Test.

Calculating the homes required

To calculate the total net homes required over the 3 year period the calculations draw on:

• Annual average household growth over a 10 year period. This is calculated for each of the test years by authority based on household projections^{[footnote 1]}. The household projections used for each test year are: 2012-based in 2015/16; 2012-based in 2016/17; and, 2014-based in 2017/183;
• Local plan information^{[footnote 2]}; both the most recent local plan and the previous local plan. Only local plans which set out an authority’s housing requirement are used. Data collected includes: the adoption date, start date, end date, housing requirements including trajectories, joint plan requirements, traveller requirements, and any unmet need which has been given to or taken from other authorities; and,
• Current London Plan annual monitoring targets 2015-2025 by borough.

Given the wide ranging status and characteristics of local plans across England, the exact process to calculate results varied. Below sets out the 5 steps taken for all authorities and then goes through individual steps taken contingent on the plan status and characteristics which affect the application of the Housing Delivery Test.

For every local planning authority, the steps below were taken:

1. For every local authority annual average household growth over 10 years is calculated based on the household projections available as at 1st April in each test year^{[footnote 1]}.
2. In some cases annual average household growth may be negative for one or more of the years being tested. In these instances all 3 years of the household growth including negative values are summed. If the 3 year summed result is a negative value, this is set to zero. If the sum of the 3 years of annual household growth is positive (despite one or more negative years), the figure is not changed.
3. Net unmet need is calculated for each authority by summing all the need taken (the authority becomes responsible for delivering this housing) and taking away all need given (the authority is no longer responsible for this housing).
4. To convert this into an annual figure, the number of years the plan covers is calculated by tak- ing the difference between the start date of the plan and the end date of the plan; classifying both the start date and end date as days the plan covers. The total net unmet need figure by authority is then divided by the total plan period.
5. The annual net unmet need for each authority is then added to the annual average household growth for each test year.

In addition to steps 1 to 5, for an authority without an ‘up to date’^{[footnote 3]} local plan the steps below were taken:

1. The housing requirement is based on the household growth plus unmet need figure, set out in an adopted plan, (as calculated in steps 1 to 5 above) for each year).
2. The number of homes required for each year over 3 year period is based on the lower of the housing requirement (as above) or household growth plus unmet need. In the incidence where a local authority has no up to date local plan, the ‘lower of’ policy does not change the total number of homes required.

In addition to steps 1 to 5, for an authority with an ‘up to date’ local plan for the entire Housing Delivery Test period the steps below were taken:

1. The housing requirement is based on the annual target from the most recent plan for each year. If the local plan includes a stepped requirement then the annual target from the stepped requirement corresponding to the relevant period is used.
2. The number of homes required each year over the 3 year period is based on the lower of the housing requirement or household growth plus unmet need in each year. For each individual year that the annual target is greater than annual average household growth plus unmet need, the lower figure is used.
3. The final number of homes required sums each year’s lower figure to calculate the 3 year total number of homes required.

In addition to steps 1 to 5, for an authority with an ‘up to date’ local plan which expired during the Housing Delivery Test period the steps below were taken:

1. The housing requirement is based on the annual target from the most recent plan for each year. If the local plan includes a stepped requirement then the annual target from the requirement corresponding to the relevant period is used, and the steps below are applied in the same way.
2. If the plan becomes out of date then from this point onwards the housing requirement is based on annual average household growth plus unmet need.
3. If the plan becomes out of date midway through a year, a weighted average for that year is calculated. This means that for as many days that the plan was ‘up to date’ within a test year, the annual target is used. For the remainder of the year, annual average household growth plus unmet need is used.
4. The number of homes required each year over the 3 year period is based on the lower of the housing requirement or household growth plus unmet need in each year.
5. In this instance, the housing requirement in a given test year could be: a. The target from the plan (if the plan is ‘up to date’ for the full year); b. A weighted average of the target from the plan and annual average household growth plus unmet need (if the plan is only ’up to date’ for part of the year); or, c. Household growth plus unmet need (if the plan is out of date for the full year).
6. For each individual year that the annual target is greater than annual average household growth plus unmet need, the lower figure is used.
7. The final number of homes required sums each year’s lower figure to calculate the 3 year total number of homes required.
8. Similarly to the above, if an authority adopts a plan which covers the latter part of the Housing Delivery Test period then prior to the period the plan covers, annual average household growth plus unmet need is used. From the point following the plan start date, the housing target from this is used (using the relevant stepped requirement where appropriate). The “lower of” policy is applied in the same way, comparing the housing requirement in each year to annual average household growth plus net unmet need.

In addition to steps 1 to 5, for an authority with a previous plan which was ’up to date’ for part of the Housing Delivery Test period and whose current plan was ‘up to date’ for part of the Housing Delivery Test period the steps below were taken:

1. If the previous plan ‘up to date’ for part of the Housing Delivery Test period, but the current plan start date encompasses the Housing Delivery Test period that the previous plan covered, the most recent plan is used and the previous plan is not.
2. If the previous plan is ‘up to date’ for part of the Housing Delivery Test period and this period does not overlap with the current plan, then the previous plan is used for as long as it is valid. After this period the most recent plan is used.
3. If there is a period where there is no plan, then annual average household growth plus net unmet need is used for this period.
4. Where the periods above do not align with a test year, a weighted average for that year is calculated. This means that for as many days that the previous plan was ‘up to date’ within a test year, the annual target from that plan is used. For the remainder of the year, either the current plan or annual average household growth plus unmet need (depending on the status of the current plan) is used.
5. The “lower of” policy for each year is applied based on the housing requirement and annual average household growth plus net unmet need. For each individual year that the annual housing requirement is greater than annual average household growth plus unmet need, the lower figure is used. The final number of homes required sums each year’s lower figure to calculate the 3 year total number of homes required.

In addition to steps 1 to 5, for a London borough including Development Corporations with full plan making and decision making powers^{[footnote 4]} the steps below were taken:

1. London boroughs are treated the same as the other authorities, except where there is no ‘up to date’ plan the most recent adopted London Plan annual borough target^{[footnote 5]} is deferred to (in the place of household growth plus net unmet need) rather than household growth for the measure of need. If a London Borough has a previous plan that was ’up to date’ for part of the Housing Delivery Test period, then this plan is used for the period it applies for.
2. If the borough plan becomes out of date midway through a year, a weighted average for that year is calculated. This means that for as many days that the plan was ‘up to date’ within a test year, the annual target is used. For the remainder of the year, the annual borough target from the London Plan is used.
3. If the borough plan includes a stepped requirement then the annual target from the requirement corresponding to the relevant period is used, and the steps above are applied in the same way.
4. The number of homes required each year over the 3 year period is based on the lower of the housing requirement or household growth plus unmet need in each year.
5. In this instance, the housing requirement in a given test year could be: a. The target from the borough plan (if the plan is ’up to date’ for the full year); b. Annual borough target from the London Plan (if the plan is not ‘up to date’ for the full year); or, c. A weighted average of the target from the borough plan and annual borough target from the London Plan (if the plan is only ‘up to date’ for part of the year); or, d. A weighted average of the target from the previous borough plan and the weighted average from the current borough plan (if the previous plan is ‘up to date’ for part of the period, and the current borough plan follows directly from this) e. A weighted average of the target from the previous borough plan and the annual borough target from the current London Plan (if the previous plan is ‘up to date’ for part of the year and there is no current borough plan to cover the rest of the test year period) f. A weighted average of the target from the previous borough plan, the borough target from the current London Plan and the target from the current borough plan (if the previous plan is ‘up to date’ for part of the period, and the current borough plan follows indirectly from this meaning the London Plan is relied upon in the interim period)
6. For each individual year that the annual housing requirement is greater than annual average household growth plus unmet need, the lower figure is used.
7. The final number of homes required sums each year’s lower figure to calculate the 3 year total number of homes required.
8. The London Legacy Development Corporation’s housing requirement is based on their DELTA return in the same way as other authorities. Given there is no separately published annual average household growth, the “lower of” policy is not applied. Therefore the target from their local plan is their final number of homes required.

In addition to steps 1 to 5, for authorities covered by a joint plan with a joint requirement the steps below were taken:

1. Where authorities have a joint plan and are being measured jointly for the purpose of the Housing Delivery Test the housing requirement, where the joint plan is ‘up to date’, the joint plan target is used.
2. If the joint plan becomes out of date during the test period then a weighted average for the year is calculated; using the joint plan target up to this point and from then onwards the total house- hold growth plus net unmet need across all of the component local authorities is deferred to.
3. For each individual year that the joint annual housing requirement is greater than the joint annual average household growth plus unmet need, the lower figure is used.
4. The final number of homes required sums each year’s lower figure to calculate the 3 year total number of homes required.

The housing requirement for travellers:

In addition to steps 1 to 5 and the relevant process depending on local plan status, where applicable the requirement for traveller accommodation^{[footnote 6]} are added to the housing requirement.

This addition takes place prior to the application of the “lower of” policy which compares the housing requirement (containing the travellers housing requirement) and the annual average household growth plus net unmet need, choosing the lower for each year as the number of homes required for that teat year.

The travellers housing requirement is calculated through the steps below:

1. The travellers accommodation requirement is based on the travellers annual target from the most recent plan for each year. If the local plan contains no travellers requirement then no ad- justment is made.
2. Where there is a travellers requirement and the plan becomes not ‘up to date’ then from this point onwards the travellers adjustment is no longer made.
3. If the plan becomes not ‘up to date’ midway through a year, then for as many days that the plan was ‘up to date’ within a test year, the travellers annual target is used (apportioned based on how many days this was ‘up to date’). For the remainder of the year, no travellers housing requirement is added.
4. If the local plan includes a stepped requirement for the travellers requirement, then the annual target from the requirement corresponding to the relevant period is used, and the steps above are applied in the same way.

Calculating the homes delivered

To calculate the total net homes delivered over the 3 year period the calculations draw on:

• Net additional dwellings^{[footnote 7]} by local authority district, England 2001-02 to 2017-18^{[footnote 8]};
• Housing supply; communal accommodation, component flows of by local authority district, England 2015-16, 2016-17 and 2017-18 (published as part of the Net supply of housing release);
• Number of students in student only household by the number of bedrooms where all students are aged 18 and over, Office for National Statistics based on 2011 census^{[footnote 9]}, ^{[footnote 10]};
• Age of Household Reference Person by number of adults in household where all household reference persons are aged 16 and over, Office for National Statistics based on 2011 census ^{[footnote 10]}, ^{[footnote 11]};
• Data supplied to the department by local planning authorities and National Parks about the number of homes within a local planning authority boundary that were delivered in a National Park, 2015-16, 2016-17 and 2017-18; and,
• Data supplied to the department by the Greater London Authority about housing completions in the London Legacy Development Corporation that was within different local authority boundaries, Residential Completions between 01/04/2017 and 31/03/2018 by Planning Authority from London Development Database.

To calculate the homes delivered, the steps taken were:

1. For each local planning authority, net additional dwellings for the test years are used as the starting point. These are then adjusted in a number of ways.
2. For each authority estimate the number of net dwellings that would be freed up from the net additional student and other communal accommodation built over the 3 year period. This is calculated by dividing each authority’s student bedspaces annual net change by the national ratio (2.5) and each authority’s other communal bedspaces annual net change by the national ratio (1.8).
3. Adjust the net additional dwellings for each authority by the annual estimate of the number of dwellings which would be required in the absence of the net communal accommodation bedspaces change.
4. For the local planning authorities whose boundaries overlap with a National Park, remove the net additional homes delivered in the National Park^{[footnote 12]} each year from the net additional dwellings statistics.
5. For the local planning authorities whose boundaries overlap with the London Legacy Development Corporation, for the periods that the local planning authority’s delivery is based on the London Plan, the net homes delivered in the London Legacy Development Corporation are removed from the net addition dwellings statistics based on the data provided to the department by the Greater London Authority.
6. Borough plans which pre-date the formation of the LLDC do not give a requirement broken down into dwellings in the borough and dwellings in the London Legacy Development Corporation; however the London plan does give separate plan numbers. This means where the borough’s requirement is based in the London Plan this will not include the homes to be delivered in the London Legacy Development Corporation however when the requirement is based on their own plan it will include these homes. Therefore the adjustment to net additional dwellings, removing those homes delivered in the London Legacy Development Corporation, should only be applied for the period where the requirement is based on the London Plan. For these periods, neither the requirement nor the net additions include the homes in the London Legacy Development Corporation.
7. If the borough’s requirement is based on different sources within a test year (for example, the borough plan and then the borough target from the London plan then the removal of net additions each year is scaled to the proportion of the year that the London Plan is relied upon for the housing requirement.

The final measure of the homes delivered is the sum of the annual net additional dwellings adjusted for National Park and Development Corporation delivery and the estimated net change in the dwelling stock due to the change in communal accommodation bedspaces.

Calculating the results

The final result for each authority is based on their number of homes required over the 3 year period and the adjusted net additions over the same period.

In all cases, the requirement for the number of homes is based on the lower of their annual housing requirement (based on the local plan status and characteristics) and annual average household growth adjusted for net unmet need. The lower figure in all years are summed to calculate the number of homes required over the total 3 year period for the purpose of the Housing Delivery Test.

The annual net additional dwellings adjusted for National Park and Development Corporation delivery and the change in communal accommodation bedspaces are summed to calculate the total number of homes delivered over the 3 years for the purpose of the Housing Delivery Test.

Comparing these 2 totals gives the individual Housing Delivery Test result for a given authority, joint plan or development corporation.

Housing Delivery Test (%) =
Total net homes delivered over 3 year period
divided by
Total number of homes required over 3 year period

In cases where the total number of homes required is zero the Housing Delivery Test result is undefined and no consequences apply.

Footnotes

1. Household projections as available at 1 April in each of the corresponding test years. Annual average household growth is calculated by calculating the total household growth between the test year and 10 years in the future, then dividing this by 10. For example: For the test year 2015/16 the 2012 based household projections are used. The total household growth over this period is calculated by taking the difference between the number of household in 2015 and 2025. This is converted into an annual average by dividing this total change by 10. ↩ ↩²

2. Local plan information was collected by the department from authorities using Delta webform. In some instances, the information as submitted was adapted for the Housing Delivery Test 2018 measurement and decisions about what information to use were decided on a case by case basis. For example, where plans were adopted following the submission of data from local authorities, information has been sourced from the published plans. ↩

3. An ‘up to date plan’ for the purposes of the Housing Delivery Test is a plan which is less than 5 years old, or is older than 5 years old been reviewed and found not to require updating. ↩

4. There is no 2018 Housing Delivery Test result for Old Oak and Park Royal Development Corporation as they have been recently created and do not have published targets against which to be measured. ↩

5. The London Plan figure remains valid for 5 years from the date of adoption, even where the plan is undergoing a revision. ↩

6. Meeting the definition in the Planning Policy for Traveller Sites (August 2015). ↩

7. As defined in the Housing Flow Reconciliation guidance. Net additions measure the absolute increase in stock between one year and the next, including other losses and gains (such as conversions, changes of use and demolitions). ↩

8. The 2017/18 published figure for Thanet has been manually adjusted by MHCLG to remove 84 units incorrectly included in the raw data by the authority. These units were empty homes returning to use and were removed as they did not meet the definition of a net additional dwelling. See net supply of housing statistics. ↩

9. Used to calculate the national average number of adult students living in a student only household where all students are aged 18 and over (2.5). ↩

10. These tables were specifically commissioned for this purpose, now published by the ONS. CT0773 - Number of students in student only household; CT0774 - Age of Household Reference Person (HRP) by number of adults in household ↩ ↩²

11. Used to calculate the national average number of adults living in a household where all Household Reference Persons are aged 16 and over (1.8). ↩

12. As provided to the department by local planning authorities and National Parks. ↩