Guidance

Forthcoming change: dependent development guidance

Updated 17 April 2025

Description: updates to dependent development guidance

Unit: A2.2 (induced investment)

Change announced: April 2025

Expected release date: May 2025

Description

This forthcoming change sets out a compendium of updates to the dependent development guidance in TAG unit A2.2. This incorporates the findings of internal research which identified areas where the current dependent development guidance could be improved.

The forthcoming changes announced in April 2025 will become definitive guidance in May 2025, alongside the forthcoming changes announced in November 2024, in line with the TAG orderly release process. We do not intend to publish forthcoming change versions of any guidance units or the TAG data book to accompany the changes announced in April 2025.

Detail

Summary of the changes

Internal research was recently conducted to identify some of the shortcomings of the current dependent development guidance as outlined in TAG unit A2.2. This research highlighted 2 issues with the current dependent development guidance. These are:

• there is scope to improve clarity in TAG regarding dependent development
• evaluation evidence from past appraisals shows a strong tendency to overestimate the scale of dependent development benefits

To address these 2 issues, we propose the following updates to the dependent development guidance in TAG unit A2.2:

• Updates to clarify the definition of dependent development. (appendix 1). This overhauls chapter 3 (quantifying and valuing dependent development), specifically section 3.1 (introduction) and introduces a new sub section 3.2 (what is and isn’t dependent development). The key changes and additions are:

  • Section 3.1: clarified definitions such that there are 2 types of dependent development: directly enabled and partially enabled. The section is also split into distinct sub sections covering definitions, land use change, reasonable level of service and economic narrative. It also includes a new table on ‘land use typology’ which categorises land use change depending on the scenario (do minimum (DM) or do something (DS)), and on additionality and displacement.

  • Section 3.2: a new section which uses illustrative examples to help practitioners define dependency. The examples are not prescriptive – to define dependency the method must still be followed as is now.

  • The rest of the chapter remains unchanged, such that section 3.3 is now ‘quantifying dependent developments’ (was 3.2).

• Update to add guidance on how uncertainty around dependent development benefits should be considered in the economic narrative (appendix 2):

  • Research into evaluation evidence showed that there was a strong tendency to overestimate dependent development benefits in transport appraisals.

  • The first of the changes we will make to the dependent development guidance to account for this bias is to add a section outlining how practitioners should address uncertainty around the delivery of dependent development benefits.

  • The new guidance will also include the following questions for practitioners to consider in the economic narrative:

    • Is dependent development a key part of the expected benefits arising from the transport investment scheme?

    • What proportion of total development is dependent? The lower the proportion of total development that is dependent the higher the risk that dependent development benefits could be overestimated.

    • Is there any expectation or evidence that local constraints (such as regulatory or land scarcity constraints) may limit or slow down the delivery of dependent development units?

• Update to add a sensitivity test to the dependent development guidance (appendix 3):

  • Research into evaluation evidence^{[footnote 1]} showed that there is a strong tendency to overestimate dependent development benefits in transport appraisals.

  • To address this, we will introduce a sensitivity test which would give practitioners a methodology to demonstrate how sensitive a scheme’s indicative benefit cost ratio (BCR) and value for money (VfM) category are to the assumptions made on dependent development delivery.

  • The sensitivity test would aim to highlight uncertainty around the central appraisal estimates and re-enforce the role of considering uncertainty in determining the final VfM category, in line with the department’s value for money framework.

  • The sensitivity test will be a switching value where practitioners will calculate the percentage decrease in dependent development benefits required to move the indicative BCR into the lower category.

  • This switching value will be compared to the average overestimate of dependent development benefits, which is calculated from the evaluation evidence.^{[footnote 2]} The average overestimate will be presented as a range to capture the differing degree that past schemes have overestimated dependent development benefits.

  • This will give the percentage decrease context, allowing practitioners to judge how likely it is that the VfM category could be impacted by an overestimation of dependent development benefits.

  • If the indicative BCR is less than 1 then a switching value can’t be applied, in this case practitioners should compare the indicative BCR with the adjusted BCR. The average overestimate of dependent development benefits, which is calculated from the evaluation evidence, will then be used to contextualise the difference between these.

Anticipated scale of change on appraisal outcomes

These changes are not expected to have a direct impact on appraisal outcomes. This is because the updates do not change the fundamental methodology for estimating dependent development benefits. The updates clarify the definition of dependent development, provide advice on how to discuss uncertainty around the scale of dependent development benefits in the economic narrative, and provide methodology on how to test the sensitivity of appraisal outcomes to dependent development delivery.

Even though the updates do not change the methodology used to estimate dependent development benefits, the sensitivity test will mean practitioners have a greater understanding of the downside risk to dependent development benefits. This may impact decision-making as more caution will be needed when interpreting estimated dependent development benefits and the resultant impact on the VfM category. The extent of the impact is difficult to predict as it will depend on the context of each scheme, which is difficult to judge.

WITA software

Dependent development benefits are not estimated in WITA so no changes to the WITA software are required.

Appendix 1: updates to clarify the definition of dependent development

This section provides extracts from the key changes and additions made in chapter 3.

Updated definition

Dependent development typically has one of these characteristics set out below:

• Development that is ‘directly enabled’ by the transport improvement. The transport improvement directly alleviates land constraints, such that the land would not have been developed at all in the absence of the intervention (i.e. the ‘without-scheme’ scenario). This is usually within the site boundary of the transport improvement – without the transport improvement the developable land does not exist or is inaccessible.

• Development that is ‘partially enabled’ by the transport improvement. This is where the transport improvement alleviates transport capacity constraints, which in turn alleviates land constraints allowing the full development to go ahead (i.e. the ‘with-scheme’ scenario). However, some of the land could have been developed in the without-scheme scenario. A dependency is likely to occur where a development will breach ‘a reasonable level of service’ on the transport network. This is a tipping point where the existing transport network cannot reasonably accommodate the additional traffic associated with the full development, hence the need to provide complementary transport investment. In this ‘partially enabled’ scenario, some of the development could take place without the transport improvement until the tipping point is reached.

In both cases, land constraints are being alleviated but in slightly different ways, and there must be a clear link to a market failure (usually ‘land rationing’). In addition, a dependency test is usually required to establish whether the development is 100% dependent (therefore ‘directly enabled’) or only partially dependent (e.g. a portion of the development can go ahead without overwhelming the transport network). However, there may be some specific cases of ‘directly enabled’ development where a dependency test is not required. Section 3.2 provides illustrative examples of how this may work in practice.

Land use typology

Dependent development is distinguished from other types of land use change, such as:

• where it is an unintended consequence of a transport intervention
• where a transport intervention is used to improve the attractiveness of an area as a place to live or work, thereby encouraging development, rather than to accommodate the additional traffic

An assessment of these other types of induced investment requires supplementary economic modelling (SEM) – see TAG unit M5.3.

This also highlights the need to categorise land use change depending on the scenario (do minimum (DM) or do something (DS))^{[footnote 3]}, and on dependency and additionality. For example, dependent development is not additional if it simply displaces development that would have happened in the DM but in a different location. And transport interventions can also increase site viability that triggers development to come forward in the DS.

This table summarises the 4 types of land use change based on these definitions. The categories are complete and mutually exclusive. The typology can be used to disaggregate land use change and identify what element is dependent. This can also support the quantification and valuation of other impacts where double counting may arise.

Table 1: land use change typology

Name	Description (including housing development example)	Is it dependent?	Is it additional? Does it add to UK housing stock relative to BAUD?	Is it part of the DM or DS, or both DM and DS?	Are the residents assumed to live elsewhere in DM?
Business as usual development (BAUD)[footnote 4]	The development that failed the dependency test[footnote 5] but is commercially viable in DM. e.g. house building that happens anyway.	No	n/a	Both DM and DS	n/a
Attracted development (AD)	Not dependent, but only viable or happens in DS. e.g. transport intervention means housing development is viable and goes ahead.	No	Yes	DS	Yes
Dependent displaced development (DDD)	Dependent development but it’s displaced from elsewhere. e.g. transport intervention unlocks land enabling housing development to go ahead. Without the transport intervention the houses would have been built in a different location.	Yes	No	Both DM and DS	Yes
Dependent additional development (DAD)	Dependent development and is additional e.g. transport intervention unlocks land which means housing development can go ahead. Without the transport intervention the houses would not have been built.	Yes	Yes	DS	Yes

What is and isn’t dependent development

Dependent developments are likely to have a wide range of characteristics such as the:

• rationale for intervention (e.g. specific market failure)
• capacity of the existing transport system,
• type of transport improvement proposed,
• scale and nature of development (e.g. housing, commercial)
• geographic spread (e.g. concentrated versus dispersed)

This means whether there is dependency is not usually clear cut. However, there are cases where some dependency is anticipated. The table below sets out illustrative examples which are based on real schemes. This is designed to help more quickly identify whether dependency is likely or not. It can also be used to sense check whether a scheme fits a particular archetype.

However, it is not prescriptive. The scheme promoter must follow the methodology to demonstrate dependency. Solely matching to an illustrative scheme in the table is not sufficient to prove dependency in of itself.

Table 2: illustrative examples

Illustrative scheme	Market failure(s) present	Is there dependent development? And is there a method to quantify and monetise this in A2.2?
(1) Office and retail development above a new railway station	Land rationing	Yes – this is an example where land is unlocked directly by the transport improvement. Without the railway station, there is no development because the land is not unlocked.
(2) A new bridge which unlocks industrial land near a port	Land rationing	Yes – this is an example where land is unlocked directly by the transport improvement. Without the bridge, the land is inaccessible and is not developed on.
(3) A large geographically concentrated housing scheme which is accompanied by public investment to improve the capacity of local roads and public transport.	Land rationing, co-ordination failure	Potentially yes (subject to a dependency test) – this is an example whereby a portion of the new development could come forward without the need to improve the transport capacity, but government investment in transport is required to enable the full housing development to go ahead. Without it the existing transport system would be overwhelmed.
(4) Mass transit scheme in a city increases attractiveness of an area, raising land values near the stations and encouraging developers to build houses in those areas rather than elsewhere.	n/a or land rationing	No – this is not a dependent development, although land rationing could mean there is a lower level of housing developing in those areas than expected.
(5) Regeneration programme that includes transport intervention(s) as well as other complimentary investments to achieve local/regional regeneration goals	Land rationing, co-ordination failure	Yes and no – depends on the nature of the scheme and market failures. It may be that some of the transport interventions may include elements of illustrative schemes (1), (2) or (3) in which case these can be quantified and monetised. But where there is no clear dependency issue based on unlocking sites or capacity, there is no method to value this. In general, it is not appropriate to use the dependent development method for very large individual and programmatic schemes that aim to have significant structural impacts on multiple, geographically dispersed, unidentified sites. An assessment of induced investment impacts for these schemes would require supplementary economic modelling.

Another way to consider dependency is through the presence of transport or non-transport complementary intervention. For example, whether it is linked to a wider policy programme such as regeneration which includes a range of non-transport interventions such as housing and commercial development. Table 3 below provides a summary of the interaction between this and dependency. In summary, the methodology set out in this unit can be used so long as there is dependency. Similar to the illustrative examples, the table below should not be used prescriptively.

Table 3: dependency and complementary interventions

	Transport led intervention	Development or non-transport led intervention
Dependency	A transport improvement that directly unlocks sites (i.e. land that could not come forward for development without the transport improvement). There is a method to quantify and monetise these benefits in this unit.	Large scale interventions which are linked to regeneration or other spatial strategies. This may be where transport unlocks a proportion of the development, whereby the existing transport system would be overwhelmed if the new development went ahead. There is a method to quantify and monetise these impacts in this unit.
No dependency	All other transport schemes which influence location choices of houses and businesses, and therefore investment choices and locations of new developments in the usual way. In this way such schemes are guiding and supporting development but without directly enabling it. Because there is no dependency there is no method to value these impacts in this unit, although any land use should ideally be captured as part of the appraisal	Large scale regeneration where there is some other issue such as low viability or coordination failure, but there is no obvious dependency issues based on directly unlocking land or where improvements in transport capacity are required. Because there is no dependency there is no methodology to value these impacts in this unit, although any land use should ideally be captured as part of the appraisal.

Appendix 2: update to add guidance on how uncertainty around dependent development benefits should be considered in the economic narrative

Below is an outline of the logic and research behind each of the 3 questions:

1. Is dependent development a key part of the expected benefits arising from the transport investment scheme?

This question is simply to identify schemes where a large proportion of overall benefits are from dependent development, as these are the schemes that should be applying the updated guidance.

2. What proportion of total development is dependent? The lower the proportion of total development that is dependent the higher the risk that dependent development benefits could be overestimated.

This question is motivated by the fact that there is a direct negative relationship between the proportion of development that is dependent, and the sensitivity to slow build out or under delivery.

To give an example, let’s say there is a scheme that will deliver 10,000 houses. In a high sensitivity scenario, 20% of the houses are dependent. This means 8,000 houses need to be built before dependent development units begin to be delivered. So, if 8,000 houses are delivered, then the dependent development benefits are zero.

However, in a low sensitivity scenario, 80% of the 10,000 houses are dependent then only 2,000 houses need to be built before dependent development benefits start to accrue. So, if 8,000 houses are delivered, there would be significant dependent development benefits, even if there is shortfall of houses built.

3. Is there any expectation or evidence that local planning constraints (such as regulatory or land scarcity constraints) may limit or slow down the delivery of dependent development units?

This question captures the potential impact local constraints may have on dependent development delivery. The question is informed by empirical research which is set out in more detail in appendix 3.

Appendix 3: update to add a sensitivity to the dependent development guidance

Rationale

Empirical research shows that the construction duration, defined as the time between shovels hitting the ground and dwellings being completed, can vary significantly depending on which area of the UK the development occurs in. For example, Ball et al find that:

A one percentage point increase in housing demand reduces the construction duration in the ‘average’ location by 2.7%, all else equal. However, the reduction in the construction duration weakens to 0.8%, 1.7%, and 2.1%, respectively, if regulatory constraints, land-scarcity related constraints, or market concentration are one standard deviation higher.

This suggests that the construction duration can vary over region depending on the strictness of the planning system, degree of land scarcity and market concentration of developers. This means that dependent development benefits could be overestimated if the construction duration is underestimated when appraising transport schemes. This is because if development takes longer than expected many of the mechanisms and feedback loops that generate the delivery of dependent development units might not occur, therefore reducing the overall scale of dependent development benefits.

Ex post evaluation found that there was a strong tendency for dependent development delivery to be significantly overestimated. For example, in the case of the Ebbsfleet development (PDF, 6MB) it was found that:

The volume of development at Ebbsfleet has significantly disappointed compared to the original plans, leading to further government initiatives to speed up delivery, including establishing the Development Corporation in 2015.

An evaluation carried out by the What Works Centre for Local Economic Growth^{[footnote 6]} looked at completed road schemes where dependent development guidance or a clear housing objective was key to justify their business case. They found that all cases had failed to meet their dependent development target, measured as number of dwellings completed by a specified date – which at the time of the research was 2017.

What Works Centre for Local Economic Growth carried out updated research in 2022 and assessed the same schemes. Table 4 shows the percentage progress towards the original expected dependent development in 2017 and 2022. The number of dwellings completed in 2017 as a proportion of the target range from 0% for the worst performing scheme up to 43% for the best performing scheme, with an average of 23% over all schemes.

Table 5 presents an extrapolation using the implied annual number of dwellings built between 2017 and 2022 as a proportion of the target, 15 years after scheme completion only the Evesham and Exeter schemes are forecast to meet their dependent development targets. The other four schemes are forecast, on average, to have only met 48% of their respective original dependent development target.

Table 4: dependent development delivery over time for each scheme

Scheme name	Scheme completion date	Original, timetable for dependent development	Original expected no. of dependent development dwellings	Total completed dwellings up to 2017	Total completed dwelling up to 2022	% progress towards original expected development – 2017	% progress towards original expected development – 2022
Turnstall northern bypass	2008	No timetable data available	395	168	292	43%	74%
East of Exeter M5 junction 29	2013	2026	20,000	3,561	11,709	18%	59%
Doncaster network Woodfield link rd	2013	No timetable data available	1,600	268	364	17%	23%
Evesham Abbey Bridge maintenance scheme	2014	2030	550	0	815	0%	148%
Newark to Widmerpool A46 improvement	2012	2026	5,000	358	1,070	7%	21%
Taunton Third Way (major scheme bid)	2011	No timetable data available	1,100	321	427	29%	39%

Table 4: Dependent development delivery over time for each scheme included in ‘Dependent development research update for Department for Transport’, What Works Centre for Local Economic Growth (What Works Growth), June 2022.

Table 5: a linear extrapolation of cumulative dependent development delivery

Scheme name	3rd year after scheme completion	5th year after scheme completion	10th year after scheme completion	15th year after scheme completion
Turnstall northern bypass	14%	24%	49%	80%
East of Exeter M5 junction 29	13%	26%	67%	107%
Doncaster network Woodfield link rd	13%	18%	24%	30%
Evesham Abbey Bridge maintenance scheme	0%	59%	100%	148%
Newark to Widmerpool A46 improvement	4%	7%	21%	36%
Taunton Third Way (major scheme bid)	15%	24%	37%	47%

Table 5: A linear extrapolation of cumulative dependent development delivery, assuming the annual rate of dependent development delivery continues at the rate it showed between 2017 and 2022 for each scheme (using the data in table 1), as a percentage of the original dependent development target presented in table 1. Based on data from ‘Dependent development research update for Department for Transport’, What Works Centre for Local Economic Growth (What Works Growth), June 2022.

An update to this research published in 2022^{[footnote 7]} found that, all the schemes had delivered additional dwellings since 2017, but only one scheme had met the original dependent development target. The update also noted that:

For 5 of the 6 schemes, the dependent development targets are unlikely to be met by the target date (or within 20 years of the scheme completion, where a target date is not given). The Abbey Bridge maintenance scheme in Evesham is the exception.

The linear extrapolation we created estimates that the Exeter scheme would meet 100% of its target in year 15 after scheme completion. The research states that all the schemes except Evesham wouldn’t meet the original target after 20 years, which suggests that the authors assumed a different rate of delivery for the Exeter scheme than we did compared to the analysis in table 5.

In conclusion, the evaluation evidence suggests there is a strong tendency to overestimate the scale of dependent development (though it should be noted that the What Works Centre for Local Economic Growth reports only looked at road schemes and delivery of housing), and geographic variation in the build out rate of housing schemes more generally. This means that dependent development benefits are likely to be overestimated which could impact VfM analysis and decision making.

To address this tendency, we propose introducing a sensitivity test into TAG that would seek to give scheme promoters the tools they need to address the tendency to overestimate. The sensitivity test will be proportionate, requiring no significant changes required to core analysis or modelling.

Switching value approach

This approach uses a switching value to show by how much dependent development delivery would need to fall for the BCR to fall enough for the transport scheme to enter a lower VfM category. This would show how sensitive the current VfM category is to estimates of the delivery of dependent development, which we know tend to be overestimated. The percentage change in dependent development delivery required to enter the lower VfM category would then be contextualised using the What Works Centre for Local Economic Growth evaluation evidence.

We have developed 2 different approaches depending on whether the indicative BCR (and equivalent VfM category) is greater or less than 1. Examples of this are set out later in this section.

To calculate the average overestimate of dependent development benefits we use the dependent development build out data from the What Works Centre for Local Economic Growth’s 2022 report. We then assume a constant rate of change in each year between 2017 and 2022 to derive an estimate of the annual number of dependent development units delivered each year between 2017 and 2022. Using that derived annual number of units delivered, we extrapolate out until the schemes exceed their original dependent development target.

Table 5 presents this linear extrapolation as a percentage of the original dependent development target. Fifteen years after schemes’ completion date, the average delivery as percentage of the original target was 75% over all schemes. If we only average across schemes that under delivered dependent development units relative to the original target, then the average delivery falls to 48%. This suggests that across all schemes the average under delivery was 25% while of the schemes that underdelivered, the average under delivery was 52%. This provides evidence supporting a contextualising range of approximately 30 to 50%, to capture the variable scale at which schemes tend to underdeliver dependent development benefits.

We then construct a counterfactual which is used to compare the extrapolation derived using the above approach. To construct this counterfactual, we utilise the data presented in table 6, which is from a 2024 report produced by Lichfields (PDF, 6MB) which includes the mean and median build out rates in dwellings per annum for differing sizes of development. Using this data, we calculated how many dependent development dwellings would be delivered by the time the above extrapolation exceeded the original target if the delivery for each scheme followed the mean build out rate.

Table 6: mean and median build out rates by site size

Site size (dwellings)	Mean build-out rate (dwellings per annum) first edition	Mean build-out rate (dwellings per annum) second edition	Mean build-out rate (dwellings per annum) third edition	Median build-out rate (dwellings per annum) second edition	Median build-out rate (dwellings per annum) third edition
50-99	27	22	20	27	18
100-499	60	55	49	54	44
500-999	70	68	67	73	68
1,000-1,499	117	107	90	88	87
1,500-1,999	129	120	110	104	104
2,000+	161	160	150	137	138

Table 6: Mean and median build out rates by site size. First, second and third edition refers to previous publications of the report. Data from Lichfields (2024) ‘Start to Finish How quickly do large-scale housing sites deliver?’ (PDF, 6MB).

The extrapolation for each scheme is then taken as a percentage of the counterfactual. This gives the delivery relative to what we would expect if delivery had followed the mean build out rate. We then average across all the schemes which gives the average delivery across all schemes for each year.

We have calculated the delivery including and excluding Exeter which was a large scheme with a dependent development target of 20,000. This is because the Lichfields paper only gives the mean build out rate for sites in select categories, with the highest site size category being in 2,000+ as shown in table 6, and it’s unclear if site sizes above this size all have similar build out rates.

Table 7: delivery of schemes compared to mean build out rate counterfactual

	Year 1	Year 5	Year 15
Turnstall northern bypass	38%	38%	80%
East of Exeter M5 junction 29	59%	69%	107%
Doncaster network Woodfield link rd	61%	52%	30%
Evesham Abbey Bridge maintenance scheme	0%	97%	152%
Newark to Widmerpool A46 improvement	48%	48%	79%
Taunton Third Way (major scheme bid)	59%	59%	47%

Table 7: Average delivery of all schemes shown in tables 1 and 2 compared to a counterfactual constructed using the mean build out rates shown in table 6. Year 1, 5 and 15 refer to the number of years since scheme completion.

Table 8: average delivery of all schemes compared to mean build out rate counterfactual

	Year 1	Year 5	Year 15
Average delivery of all schemes	44%	61%	83%
Average delivery of all schemes (excluding Exeter)	41%	59%	78%

Table 8: Average of delivery compared to a counterfactual constructed using the mean build out rates shown in table 6. Year 1, 5 and 15 refer to the number of years since scheme completion. Averages are across all schemes for the indicated year.

Table 9: fifteen year average of delivery compared to mean build out rate counterfactual

	Delivery relative to a counterfactual – 15-year average
Average delivery of all schemes	66%
Average delivery of all schemes (excluding Exeter)	63%

Table 9: 15-year average of delivery compared to a counterfactual constructed using the mean build out rates shown in table 6. The average is an average of the average of all schemes (which is shown for year 1, 5 and 15 in table 5) over each year of the first 15 years after scheme completion, rather than the average of all schemes in only year 15 shown in table 5.

Table 7 shows a sample of the average delivery relative to what we would expect if delivery had followed the mean build out rate. Table 8 shows, for illustrative purposes, the average delivery across all schemes for 1, 5 and 15 years after scheme completion.

Table 9 then presents the all-scheme average averaged over the first 15 years after scheme completion. These figures indicate that the under delivery is approximately 30 to 40% (100 minus 63 or 66). This supports the contextualising range of 30 to 50% suggested above.

Key assumptions and caveats

There are 3 key assumptions and caveats to this approach that need to be highlighted.

1. The sensitivity is only applied to the overall dependent development benefit, it does not affect the land use change assumed in the rest of the economic case.

2. It assumes all components of total dependent development benefits (i.e. land value uplift, transport external costs, land amenity value and non-transport complementary interventions) scale down linearly. So, if fewer developments are completed the components would change proportionately. This means that the sensitivity can be applied to total benefits without having to account for non-linearities in how individual components may change. It also means the evaluation data on the delivery of dependent development units can be used as a proxy for the overestimation of benefits.

3. Because the sensitivity is applied to the total dependent development benefits, no specific assumption is made about the cause of the decreased benefits. It could be caused by a slow construction rate, housing not being delivered in full or some combination of both.

Example of the switching value approach

Example 1 – indicative BCR and equivalent VfM category greater than 1

• Assume in this example that the indicative BCR is greater than 1 and the VfM category is medium (meaning a BCR greater than or equal to 1.5 and less than 2.0).
• Let’s also assume that for the BCR to fall below 1.5, and therefore the scheme to fall into the low VfM category, the estimated dependent development benefits need to fall by 25%.
• The 25% fall can then be compared to the average under delivery of dependent development benefits calculated above. Since we know that the average under delivery is up to 50%, we also know that a fall of 25% means there is a high risk that dependent development will under deliver sufficiently to affect the VfM of the scheme.
• Therefore, the uncertainty around the delivery of dependent development benefits, and the risk this poses to the VfM category, is made explicit.

Example 2 – indicative BCR and equivalent VfM category is less than 1

• If the indicative BCR is less than 1 then the switching value outlined in example 1 cannot be used, as the scheme is likely to be categorised as ‘poor’ VfM. In this scenario the indicative BCR can be compared to the adjusted BCR to illustrate the sensitivity of the analysis to the estimates of dependent development delivery.
• Let’s assume that a scheme has an adjusted BCR of 0.5 but an indicative BCR of 0.9, so the level 3 benefits are pushing the scheme close to the next VfM category.
• The indicative BCR can then be compared to the adjusted BCR, using the average under delivery of dependent development to judge how likely it is the indicative BCR will be an accurate reflection of a scheme’s costs and benefits.
• For example, if reducing the dependent development benefits by 50%, the average under delivery calculated above, only leads to an indicative BCR of 0.6, the indicative BCR of 0.9 might be considered to be highly sensitive to dependent development benefits being overestimated. This would need to be factored into the final VfM decision which considers other non-monetised benefits and uncertainty.

Contact

For further information on this guidance update, please contact:

Transport Appraisal and Strategic Modelling (TASM) division
Department for Transport
Zone 2/25 Great Minster House
33 Horseferry Road
London
SW1P 4DR

tasm@dft.gov.uk

1. See annex 3 for further details of how the average overestimate was calculated. ↩

2. See the footnotes in annex 3 for the specific sources. ↩

3. DM and DS are equivalent to ‘without-scheme’ and ‘with-scheme’ scenarios respectively. ↩

4. As in the National Trip End Model (NTEM) which contains forecasts of population, households and employment and allocates this across Great Britain. ↩

5. See subsequent sections for more information on the dependency test. ↩

6. ‘Evaluating the Performance of Dependent Development’, What Works Centre for Local Economic Growth, Lynne Miles and Patrick Andison, April 2018, ARUP ↩

7. ‘Dependent development research update for Department for Transport’, What Works Centre for Local Economic Growth (What Works Growth), June 2022 ↩