Additionality Guide · (referred to as ‘The Green Book ... in any detail in this Guide.7 English Partnerships | Additionality Guide Third Edition| October 2008 Introduction 1 Introduction

x M] – [Gl*x (1-L*) x (1-Dp*) x (1-S*) x M*]

Al= [Gl x (1-L) x (1-Dp) x (1-S) x M] – [Gl*x (1-L*) x (1-Dp*) x (1-S*) x M*]

A l= [Gl x (1-L) x (1-Dp) x (1-S)




Al= [Gl x (1-L) x (1-Dp) x (1

x (1-Dp) x (1-S) x M] – [Gl*x (1-L*) x (1-Dp*) x (1-S*) x M*]


x (1-S) x M] – [Gl*x (1-L*) x (1Dp*) x (1-S*) x M*]


Additionality GuideA standard approach to assessing

the additional impact of interventions

Additional impacts

The reference case (deadweight)

Adjusting the reference case and intervention options

Calculating additionality

Method StatementThird Edition

1 Introduction 1

2 Additional impacts 3

2.1 Overview 3

2.2 Assessing the additionality of outputs and outcomes 7

2.3 Relevance by intervention type 8

2.4 Target area or group 9

2.5 Time period for the appraisal 10

2.6 Quality of outputs and outcomes 11

2.7 Maximising additionality 12

3 The reference case (adjusting deadweight) 13

3.1 Overview 13

3.2 Measuring the baseline position 13

3.3 Assessing the reference case 13

3.4 Evidence from evaluations and research 15

4 Adjusting the reference case and intervention options 17

4.1 How to assess additionality of each option – the factors explained 17

4.2 Leakage 17

4.3 Displacement 21

4.4 Substitution 23

4.5 Economic multiplier effects 24

5 Calculating additionality 27

5.1 Introduction 27

5.2 How to calculate additionality 27

5.3 Illustrative worked examples by intervention type 28

5.4 A worked example from the evidence base 38

5.5 Gross to net additionality ratios 38

6 Conclusion 41

Glossary 43

Bibliography 44

Appendix A 45

Appendix B 46

Appendix C 48

Appendix D 49

Appendix E 50

Contents

Contents

This Guide explains how toassess the additional impact oradditionality of a regeneration,renewal and regionaldevelopment intervention1.Additionality is the extent towhich something happens as aresult of an intervention thatwould not have occurred in theabsence of the intervention.

It is the third edition2 of the Guide, whichhas been updated to include additionalinformation on how to assess additionality inrelation to the qualitative effects of publicsector interventions, such as meeting highersustainability standards. In addition, itincludes recent evidence from evaluationsabout the actual scale of additionality andthe various factors that need to beconsidered. Also a number of the exampleshave been revised and new ones added tobetter illustrate how to assess additionality.Further details of the main changes sincethe last edition of this Guide are set out atAppendix A.

The approach to assessing additionalityremains consistent with:

HM Treasury’s Guide to Appraisal andEvaluation in Central Government(referred to as ‘The Green Book’3); and

Guidance on Assessing the Impacts ofSpatial Interventions (referred to as the3Rs4 guidance) produced by the, then,Office of the Deputy Prime Minister(ODPM).

Project appraisal entails being clear aboutobjectives, thinking about alternative waysor options of intervening to meet them,estimating and presenting the costs andbenefits of each potentially worthwhileoption and taking full account of associatedrisks. It is an important management tooland is essential to good decision-makingbecause it:

Provides information to redesigninterventions in order to maximise theirimpact;

Helps to test ideas and selectinterventions that will work;

Enables decision-makers to make thebest possible decisions; and

Produces more effective and efficientinterventions that deliver real results.

Central to good appraisal is the need toassess whether the intervention concernedwill bring additional benefits over and abovewhat would have happened anyway in itsabsence.

However, assessing the additional outputs5

and, where possible, outcomes6 of anintervention option is only one of the stepsinvolved in appraising an intervention. ThisGuide is primarily concerned with themethodology for calculating additionality.There are many appraisal issues that affectthe ability to measure additionalityaccurately, such as defining options,measuring outputs/outcomes and assessingthe risk associated with each option. Theseissues are touched upon but not dealt within any detail in this Guide.7

English Partnerships | Additionality Guide Third Edition | October 2008

Introduction

1 Introduction1

Introduction1

1 An intervention is defined as being any project, programme or policy that is implemented or supported by the publicsector in order to achieve its objectives.

2 A first edition of this paper was prepared by AMION Consulting in association with Dr Peter Tyler in May 2001 – English Partnerships, A Standard Approach to Assessing the Additional Impacts of Projects – Method Statement. The second edition was produced by AMION Consulting in September 2004.

3 H M Treasury (January 2003) – Appraisal and Evaluation in Central Government. The Green Book is a high leveldiscussion of the principles and best practice covering all issues related to project appraisal.

4 The, then, ODPM (2003) – Assessing the Impact of Spatial Interventions: Regeneration, Renewal and RegionalDevelopment.

5 Outputs are the things immediately produced by a project, such as land reclaimed or number of trainees.6 Outcomes are the intended results of the project in terms of its key or ultimate objectives, such as sustainableincreases in economic activity.

7 Further information on these issues can be found on the Office of Project Advice and Training (OffPAT) website –www.offpat.info

Additionality is relevant to various stages ofan intervention’s lifecycle, including strategydevelopment, appraisal, monitoring andevaluation. However, whilst the samethought processes and logic applies toeach stage, the focus of this Guide ison appraisal.

The resources allocated to assessing theadditional impact of an intervention shouldbe proportionate to the nature and scale ofthe intervention. Interventions that are novel,contentious, repercussive or involve a highlevel of risk will require more in-depthanalysis, as will larger interventions – inother words, those that involve a significantamount of public expenditure. Standardinterventions of a type that have beendelivered before will typically require lessoriginal analysis and can more easily drawon appraisals and evaluations of previousinterventions. For smaller interventions, it willoften not be feasible to undertake anydetailed assessment of additionality. In thesecases, appraisal should be informed by aqualitative assessment and statement aboutintervention additionality, as well asprogramme and policy level assessments.

It is recognised that assessing additionalityis not an easy task and that generally thiswill be carried out by specialists or thosewith experience in intervention developmentand appraisal. This Guide is generally aimedat economists and other suitably qualifiedand experienced professionals. However, itis also intended to be accessible to non-specialists in order to provide anunderstanding of the principles andimportance of assessing additionality. TheGuide does cover some material of a moretechnical nature and where it is necessary touse technical terms we have sought toexplain each term or concept carefully andprovide examples.

The Guide continues in the following fivechapters:

Chapter 2: sets out the basicmethodology and key issues in order toassess the additional impacts of anintervention;

Chapter 3: discusses the reference caseor deadweight – in other words whatwould happen anyway, without theintervention;

Chapter 4: explains the adjustments thatneed to be made to the the interventionoptions and reference case to calculateadditionality;

Chapter 5: presents examples of how tocalculate additionality; and

Chapter 6: sets out a number ofconcluding remarks.


Introduction

Introduction

12

2.1 OverviewMost interventions will have bothpositive and negative impacts. Inappraising the effects of anintervention it is important that allof these are taken into account inorder to assess its additionalimpact or additionality – in otherwords, the net changes that arebrought about over and abovewhat would take place anyway.

H.M. Treasury’s Green Book states that animpact arising from an intervention isadditional if it would not have occurred inthe absence of the intervention. The, then,ODPM's guidance on Assessing theImpacts of Spatial Interventions definesadditionality as:

‘The extent to which activity takes placeat all, on a larger scale, earlier or within aspecific designated area or target groupas a result of the intervention’’

In addition, greater quality can provideadditional effects. The additional impact ofan intervention is therefore the differencebetween the reference case position (whatwould happen anyway) and the position ifthe intervention option was implemented(See Figure 2.1).

Additionality may relate to:

Scale – a greater quantity of houses orjobs may be delivered in an area.

Timing – activity may happen earlier thanwould otherwise have been the case.Where there are very large timing

differences, it may be appropriate todiscount the profiles of quantified outputsand outcomes. However, this is only likelyto be an issue for a small number ofinterventions. Further information on thiscan be found in Annex 7 of the 3Rsguidance.

Specific area or group – the extent towhich the target beneficiaries actuallybenefit from an intervention. For example,for a key worker housing project, will all ofthe occupiers be key workers?

and/or

Quality – the quality of the outputs /outcomes may be different because of apublic sector intervention. (Further detailsabout how to assess additional qualitativeimpacts are set out in Section 2.6).

It is possible to assess the additionalimpacts of an intervention using twoalternative approaches, as follows:

Top-down – by assessing expectedchanges in overall indicators, such as thelevel of employment, total population ornumber of dwellings (this is alsosometimes referred to as the outcomeindicator approach).

and/or

Bottom-up – appraising the expectedimpact of individual actions or projects,through consideration of their likelyoutputs and outcomes.

At the strategy or programme developmentstage the focus will often be, albeit notexclusively, on top-down approaches, whileat the intervention design and developmentstages, the principal emphasis is on bottom-up analyses. Both approaches are oftenused in evaluations.


Additional impacts

2 Additional impacts3

Additionalimpacts

2

Figure 2.1: Net additionality framework

Less

Impact ofintervention

option

Impact ofreference case(deadweight)

net additionalimpact=

Figure 2.2 across shows an example of top-down analysis of expected overallemployment change in an area as a result ofan intervention. It shows that, over a ten-year period, total employment in an area isexpected to rise from 8,000 to 20,000.However, without the intervention, it isestimated that employment would havegrown to 15,000 anyway. Thus, whilst thegross change in employment is 12,000, thenet additional impact of the intervention is5,000 jobs.

Alternatively, if employment in the area wereexpected to decline to say 5,000 jobs if theintervention did not go ahead then the netadditional impact at the end of the appraisalperiod would be 15,000 jobs (i.e. thedifference between 5,000 jobs and 20,000jobs). Further examples of various possiblerelationships between the intervention optionand reference case are set out at Appendix B.

However, in reality it will not normally bepossible to appraise interventions bydynamically modelling net effects at aspecific target level. In the vast majority ofcases there are serious difficulties in usingsuch top-down approaches to assess whatare usually highly marginal gross and netimpacts. Therefore, the focus in appraisal isupon a bottom-up or intervention specificapproach to assessing additionality. In orderto do this a clear analytical framework isneeded.

To calculate the total net additional localeffects of an intervention, a number ofadjustments need to be made to both theintervention and reference case options. Theapproach adopted to assessing theadditionality of an intervention is shown inFigure 2.3.

The figure introduces a number of importantkey concepts, as set out in Box 2.1.

For very large interventions, considerationwill also need to be given to crowding outand crowding in effects. These effectsoccur where increases in public expenditurecause other variables in the economy toadjust resulting in either a decline (crowding

out) or increase (crowding in) in privateexpenditure. These effects are normallyconsidered only in very large appraisalsconcerned with regional and national levelimpacts. Further information on crowdingout and crowding in effects is given atAppendix D.

It is important to recognise that theanalytical framework presented above doeshave a number of limitations, in particular inaccounting for macro-economicadjustments, which may reduce (or increase)the additionality of an intervention at widerspatial scales. This limits its appropriateness


Additional impacts

Additional impacts

24

Box 2.1: Key components of additionality

Intervention options: the alternative ways or options that the public sector mightchoose in order to intervene to achieve its objectives. An estimate will need to be madeof the level of target outputs/outcomes that would be produced under each of thealternative intervention ‘do something’ options. Further discussion of the generation ofoptions is presented in Appendix C.

Reference case: the estimate of what level of target outputs/outcomes would beproduced if the intervention did not go ahead. It is the ‘do nothing’ or do minimumoption and the outputs/outcomes produced under this option are referred to asdeadweight. In some cases, deadweight might be estimated by assuming that aproportion of the total gross additional local effects would go ahead anyway under thereference case. For example, if it were estimated that 25% of, say, 200 total net localtraining places would have been provided anyway, then some 50 total net local trainingplaces would be deadweight and the total net additional local would be 150. However,the preferred approach is to construct and quantify a specific reference case scenario.

Gross direct effects: an estimate of the total effect of an intervention option or thereference case in terms of a specific output. This would mean consideration of widerconsequential or induced effects, as well as the immediate effects8. It may also berelevant to consider agglomeration effects. This would include, for example, the positiveeffect of a housing scheme on adjacent property values or in causing private sectorhousing development to take place which would, for example, not otherwise havehappened or would have happened later. As such it is important to consider and includepositive externalities associated with the intervention.

Leakage effects: the number or proportion of outputs (occurring under the referencecase and the intervention options) that benefit those outside of the intervention’s targetarea or group should be deducted from the gross direct effects. However, it is alsoimportant in an appraisal to consider positive and negative effects on other areas.Impacts outside the target area or group should not be ignored, particularly those inother priority area or groups.

Displacement: the number or proportion of intervention outputs (occurring under thereference case and the intervention options) accounted for by reduced outputselsewhere in the target area should also be deducted.

Substitution effects: this effect arises where a firm substitutes one activity for a similarone (such as recruiting a jobless person while another employee loses a job) to takeadvantage of public sector assistance. Again these effects need to be deducted.

Economic multiplier effects: further economic activity (jobs, expenditure or income)associated with additional local income, local supplier purchases and longer termdevelopment effects then need to be added.

8 Agglomeration effects refer to the benefits firms obtain when they locate near to each other and are related to the concepts of economies of scale and network effects.


Additional impacts

Additional impacts

25

Figure 2.2: Assessing additionality (improving reference case)

Output/outcome

End ofappraisal period

Baseline

20,000

15,000

8,000

Jobs

with intervention

without intervention reference case

gros

sim

pact

net

addi

tiona

lim

pact

Gross impact = 20,000 - 8,000 = 12,000Net additionalimpact = 20,000 - 15,000 = 5,000

2008 Time 2018

Figure 2.3: Approach to assessing project level additionality – key components

Intervention options

Gross direct effects

Less leakage fromtarget area/group

Gross local direct effects

Less displacement(factor and product

market) / submission

Net local direct effects

Plus multipliereffects

Total net local effects

Gross direct effects

Less leakage fromtarget area/group

Gross local direct effects

Less displacement(factor and product

market) / submission

Net local direct effects

Plus multipliereffects

Total net local effects Total net additional local effects

Reference case (deadweight) Net additional impact

Less =

in discussing national economic efficiencyimpacts, unless it is combined with detailedmacro-economic modelling.

Since it is not practicable, or indeed oftenpossible, in the context of relatively smallinterventions to develop dynamic, fullequilibrium models, the approach proposedis a form of partial equilibrium analysis. Assuch, it is a simplification and it is essentialto ensure that the analysis is a sufficientlyreasonable representation of reality. In acontext of tighter labour and other markets,it may be important to assess pricefeedback effects and other adjustmentmechanisms, such as changes inparticipation rates and migration flows.These effects can, in principle, be handledvia the displacement adjustments although,in practice, this has often not occurred.

If prices are likely to change significantly asa result of an intervention, the analysisbecomes more complicated. For example,when considering the additionality ofeconomic activity associated with acommercial development, the followingissues need to be considered:

How far the individual development isadditional;

How far it represents a net addition to the stock of property taking account ofeffects in raising land prices or depressingrentals and the consequent effects of thison private sector development activity;and

How far the activity generated by thegreater availability and/or reduced cost ofproperty is additional.

Where an intervention results in effects ‘offthe direct causal chain’ the nature andadditionality of these effects also need to beconsidered. For example, a project toprovide homes or jobs for a particular targetgroup might be judged a failure based on anarrow view of leakage but it may have hadsignificant secondary benefits through‘freeing up’ homes or jobs which are takenup by the target group. It may be thatdifferent proportions for factors such asdisplacement need to be applied to eachimpact where the effects are not directlyassociated with the intervention.

The assessment of additionality is not amechanistic process but depends on theappraiser’s judgement and knowledge of theintervention and the wider environment. It isimportant that these judgements areinformed by evidence and the reasons forthem explained. In order to assess the levelof additionality it is necessary to considerhow the intervention has affected marketactivity, as well as its impact on other publicsector activities.

The Guide presents various estimates forthe potential scale of each of the factorsdiscussed above. However, significant careneeds to be taken in using default or readyreckoner values. The following best practiceframework for the use of these and othervalues is therefore suggested:

Best – bespoke investigation usingvarious data capture methods, such assurveys or the results of bespokeeconomic or other modelling.

Good – values chosen through a reviewof previous evaluations recognisingdifferences in:

(i) the policy and location (e.g.geographic, demographic or economicdifferences);

(ii) the assumptions made in the originalevaluation; and

(iii) significant changes in situation (due totime of investigation).

Adequate – default values chosen fromavailable guides, where the choice hasbeen carefully considered and thereasoning explained.

Not adequate – default values withoutconsideration of any of the above. Valuesused without reference to origin or fitnessfor purpose.


Additional impacts

Additional Impacts

26

2.2 Assessing theadditionality ofoutputs andoutcomes

Additionality is relevant to all outputs andoutcomes. The aim of public sectorinterventions is to achieve desiredoutcomes. However, many outcomes (suchas changing behaviours and attitudes) aredifficult to measure and will often only occursometime after an intervention has beenimplemented.

Thus, whilst interventions are concernedwith delivering outcomes, the focus ofattention in appraisals is often on assessingthe net additionality of outputs.

It is important to identify interventionoutputs9 that are expected to contribute toan improvement in the desired outcomesand for these outputs, where relevant, toconsider leakage, displacement,

substitution, multipliers and deadweight.However, it must be recognised that outputsare necessary but not sufficient to deliverdesired outcomes. Table 2.1 showsexamples taken from the New Deal forCommunity (NDC) Guidance that illustratesrelated outputs and outcomes byintervention type.

It is important that the outputs andoutcomes are accurately estimated. Anumber of sources can be used to assessthe scale and timing of outputs andoutcomes, including:

Project specific information, such as abusiness plan;

Comparable interventions;

Consultation with experts;

Standard ratios, such as, floorspace toemployment density by use;10

Specific research; and

Previous evaluation results.

Care should be taken to ensure that there isno double counting between the output andoutcome indicators used. For example, if thenumber of jobs created has been weightedto allow for the wider benefits arising fromthe high skill level attached to these jobs,this should not be added to a regionalGross Value Added (GVA) measure thatalready takes into account the same benefit.

H.M. Treasury has published work on thetendency for appraisers to be overlyoptimistic (referred to as optimism bias) andto redress this tendency appraisers shouldmake explicit, empirically based adjustmentsto the estimates of an intervention’s costs,benefits and duration. Further guidance onoptimism bias can be found in Section 5 ofthe Green Book and in CLG’s optimism biasguidance note.11


Additional impacts

Additional impacts

27

9 3Rs differentiates between activities, which are the thing that is engaged in in order to deliver outputs, such as the process of constructing a house, and outputs. We are concernedwith outputs, which are the results of that activity, such as the number of houses constructed.

10 See, for example, English Partnerships (July 2001), Employment Densities – A Full Guide11 CLG (2007) – Adjusting for Optimism Bias in Regeneration Projects and Programmes: A Guidance Note.

Table 2.1: Examples of related outputs and outcomes

Examples of Outputs Examples of Outcomes

Housing Units refurbished, demolished, built Improved housing conditions, changes in tenure,reduction of turnover, satisfaction withaccommodation

Crime Number of locks and entry phone systems Reduction in volume of crime, reducedinstalled,hours spent by a community police fear of crime relative to other areasofficer inschools, CCTV points installed

Health Number of home visits by health outreach Improved mortality rates, lower illness ratesworkers,new community health centre relative to other areas

Education Number taking part in parent-school initiative, Improved school attainment levels, improved number of contacts with truants, improvement school leaver destinations into employmentto school facilities and HE/FE

2.3 Relevance byintervention type

All of the key concepts described in Figure2.2 are relevant when it comes toconsidering the employment or economicoutputs generated by an intervention.However, not all interventions are designedsolely to generate employment or economic

outputs and outcomes. Interventions varyconsiderably in the outputs and outcomesthey produce depending on whether theyare targeted on regeneration, renewal orregional development issues relating tohousing, crime, health, education or a widerange of other service areas. Whilst the netadditionality of all intervention outputs andoutcomes should be considered, it shouldbe borne in mind that the applicability of the

key concepts depends on the interventiontype and category, as well as the individualintervention itself. Table 2.2 below examinesfor a range of different interventions when itis likely that each of the key additionalityconcepts (leakage, displacement,substitution, multipliers and deadweight)discussed previously may need to beaddressed.


Additional impacts

Additional impacts

28

Table 2.2: Assessing additionality of outputs by intervention type

Intervention type Intervention category Leakage Displacement Substitution Multiplier effects Reference case(deadweight)

Housing Social economic Always need to assess May be displacement May be relevant May exist depending Always need to assesswithin housing market on local labour and

materials used inconstruction.Generally not likely tobe relevant except aspart of long-termdevelopmentmultiplier effects

Commercial Economic Always need to assess May be displacement May be relevant Relevant to Always need to assessdevelopment within commercial employment,

property market investment and incomeassociated withconstruction phase andoccupation of thepremises.

Transport Economic Always need to assess May, for example, affect Not likely to be relevant May exist depending Always need to assessenvironmental modal split on local labour and

materials used inconstruction. Alsorelevant to induced effects.

Environmental Environmental Always need to assess Not likely to be relevant Not likely to be relevant May exist depending on Always need to assessimprovement local labour and materials

used. Generally, notlikely to be relevant

Business support Economic Always need to assess Displacement may arise May be relevant Relevant to employment, Always need to assessin relation to private investment and incomesector business support(e.g. banks) and productand factor marketdisplacement

Community & social Social Always need to assess May be relevant Not likely to be relevant May be relevant, depends Always need to assessif local labour andmaterials used

Crime prevention & Social Always need to assess Not likely to be relevant, Not likely to be relevant May be relevant Always need to assesscommunity safety/health although crime may be depending on local

shifted elsewhere. This labour and materialsis a different form of useddisplacement, whichshould be described and,where practicable,quantified in an appraisal

Training Economic Always need to assess May result in displacement May be relevant- May be relevant Always need to assessof other provision substitution of labour

2.4 Target areaor group

The level of additional outputs andoutcomes of an intervention will depend, inpart, on the area or focus of analysis. Thus,for example, the level of economic linkages– and therefore multiplier effects – will begreater if a larger spatial area is beingconsidered. As a result, when appraising anintervention the spatial level and targetgroup within that area against which theintervention is being assessed must beclearly specified. They should relate directlyto the identified need (the rationale forintervention) and be wide enough to takeinto account spill over or unintended effectson other groups, areas or markets.

It is quite common in appraisals to considereffects at different spatial levels. The fourmost often used geographical levels are:site, local/sub regional, regional andnational. However, in other cases a specificpolicy priority area may be the appropriatelevel. Many appraisals, unless taskedotherwise, concentrate on the site andlocal/sub-regional level. The appropriatetarget area should be assessed by thinkingabout the rationale and objectives of theintervention, the policy context and who isexpected to be affected. For interventionslike a website or portal, the target area ofbenefit could be region-wide or national.

The issues associated with each of the fourspatial levels are as follows:

Site

The site level is the immediate vicinity of theintervention. Very few regenerationinterventions should be assessed only at thesite level, as it is rare that the costs andbenefits will be concentrated in the actualarea of physical activity. Even for very smallinterventions, for example, a scheme toreclaim a fly tipping area for recreational use,the benefits will be much wider than just theimproved visual amenity of the site itself.The area of benefit could be within, say, aone-mile radius of the site if this is thedistance users will travel to make use of thefacility.

Local/sub-regional

The impact at the local level would beassessed to determine the effect on localpopulations or markets. For interventionsthat are expected to have sub-regionalimpacts the intervention may, for example,need to be appraised at the single ormultiple local authority level depending onthe spread of the costs and benefits.

The local level for interventions that generateemployment effects or other economicbenefits is often considered to be within therelevant travel to work area or if this is notappropriate then a 10-15 mile radius of thesite concerned if it is a physicaldevelopment. The precise delineation willdepend on the density of the settlementpattern in relation to the location of peopleand business activity. For very rural areas itis usually more. For very urban areas itmight well be less.

Regional

For interventions which are intended to haveregional impacts this is the level at which thecosts and benefits should be measured.These will typically be very largeinterventions, which generate outputs thatsignificantly add to the stock of these typesof outputs at a regional level. Given theseregional impacts it would also be necessaryto assess the impact of the intervention at asub-regional and possibly even local level asthe impacts (costs and benefits) maypositively or adversely affect localpopulations or markets.

National

Few regeneration interventions are likely tobe considered at the national level as thecosts and benefits of an individual projectare unlikely to be of such quantity that theywould change the position of a target groupor market on such a large scale or widegeographical area. An intervention wouldhave to deliver tens of thousands of homesor jobs and/or millions of square metres ofcommercial floor space before its impact atthe national level would have to beassessed.

When choosing the target group and spatialarea for appraisal it may appear cheaper

and easier to use people classifications orgeographic areas for which there is existingdata. However, if this does not accuratelyreflect the specific intervention target groupor area then the measurement of theintervention’s outputs and outcomes both atthe appraisal and the evaluation stage willbe made more difficult, if not impossible. Forexample, for an intervention designed toimprove housing conditions at the very locallevel of a large housing estate, it may betempting to use existing local authority dataon housing satisfaction at the district level toact as a baseline. Then subsequent years’data could be used for monitoring purposesbecause it is cheap and readily available.However, it will not enable a real measure ofthe outputs and outcomes of theintervention to be assessed because themore local experience, which could bebetter or worse, may not be discernible atthe district level. It is often better tosupplement existing data by further analysisor carry out additional surveys to obtaininformation that directly corresponds to thetarget group or target area.

Having thought through and clearly identifiedthe target area and group, when measuringadditionality it is worth noting that the levelof displacement and size of the multipliereffect are likely to vary with the size of thearea under assessment. The larger the areaover which the benefits of the programmeare being analysed, generally the higher will be:

The level of displacement: because thereis likely to be a greater number ofenterprises/organisations providing theproduct or service that the intervention isseeking to provide and with which it willbe competing. A high level ofdisplacement will reduce the number ofadditional outputs/outcomes.

The size of the multiplier effects: these arelikely to be greater as there are moreopportunities for economic linkages interms of suppliers and local expenditurethan there are in a smaller geographicalarea. Larger multiplier effects will increasethe number of additionaloutputs/outcomes.


Additional impacts

Additional impacts

29

The level of leakage from a target area willoften be lower as the target area grows,since there are likely to be a greater numberof target beneficiaries able to access thetarget outputs. It is also worth noting thatinterventions located on the boundary of anarea designed to tackle spatial areas orgroups are likely to suffer from higher levelsof leakage than the same interventions morecentrally located. This is becauseintervention benefits are likely to be moreeasily accessed by those in adjacentlocations. As noted above, carefulconsideration needs to be given to what isthe appropriate target area and group. A lowlevel of leakage is desirable as, all thingsbeing equal, it will increase additionality andgive better value for money. Moreover, it isimportant that each additionality factorshould be calculated using the same targetarea or group.

2.5 Time period forthe appraisal

The time period over which you areappraising the intervention should be set outand the reason for selecting it made clear.Normally the period chosen should besufficiently distant to include all importantcosts and benefits. For physical assets it willusually extend to cover its useful lifetime.The residual value of any asset at the end ofthe appraisal period would need to beincluded in the appraisal.

Often the achievement of desired outcomeswill occur only after many years following theimplementation of the intervention. Outputs,on the other hand, will normally beproduced earlier on. For example,reclamation of a brownfield site for housingdevelopment could generate five hectares ofreclaimed land in years one and two of theintervention and 100 housing units and10,000 sq m of commercial floor space inyears three and four with no further outputsfrom year five onwards. However, theoutcome of a thriving community may notoccur until, say, year ten or later.

In assessing the effect of an intervention oneof the issues that will need to be consideredis whether activity is likely to happen soonerthan would otherwise be the case. For

example, the intervention of the publicsector may enable a project to beimplemented at an earlier date. As a result,the intervention option may not onlygenerate additional outputs compared to thereference case, but also bring forward theachievement of the target outputs. In somecases, interventions may principally beconcerned with accelerating the delivery ofoutputs and outcomes.

To take account of time additionality,discount factors can be applied. Discountinginvolves reducing future costs and benefitsto reflect the fact that society places greatervalue on costs and benefits that arisesooner rather than later. This provides apresent value (i.e. the value that is placed onthem today) of costs and benefits. Inpractice only interventions with acost/benefit profile that extends over anumber of years (say three or more) areusually subject to discounting.

In addition, the impacts of an intervention orpolicy will often change over time. Take ahousing clearance and redevelopmentproject. During the early period there maybe negative effects, due to demolition andthe relocation of residents, although therecould be employment opportunities in theconstruction industry. The main outputs andoutcomes associated with new homes andcommunities will only occur later on.


Additional impacts

Additional impacts

210

2.6 Quality of outputsand outcomes

The traditional emphasis in manyassessments of additionality has been onquantitative indicators, such as the numberof housing units produced by anintervention. However, in many cases thesequantitative measures will not take sufficientaccount of the qualitative differencesbetween intervention options and thereference case. In addition, an interventionmay deliver higher quality outputs and thisneeds to be reflected in the additionalityassessment. Therefore, it is important thatthe qualitative effects are also assessed.

There are three main approaches toassessing additionality in qualitative terms:

(i) minimum thresholds – in other words toonly count gross direct outputs thatexceed a minimum quality standard.Thus, for example, only housing unitsthat meet decent homes standards or,alternatively, a specified Code forSustainable Home level may be countedas an output. Public sector support maybe offered to assist the delivery ofenhanced design or environmentalstandards. Say 100 units would bedelivered to this standard under theintervention option and none under thereference case. On this basis, the grossdirect housing units would be 100 underthe intervention option and zero underthe reference case.

(ii) weighting the outputs/outcomes througha scoring assessment – wheredifferences in the quality of outputs ariseit may be appropriate to weight theoutputs according to a scale, whichreflects the quality aspects. Qualityadjusted net additional outputs can thenbe more easily compared acrossoptions. As the quality of the outputssought will normally be affected by aconsideration of what type of outcomesare needed in an area it is not possibleto propose a standard scale. Forexample, the public sector sometimeshas the opportunity to bring on boardthe private sector at different stages ofan intervention’s development and thusgenerate greater or lesser amounts ofprivate sector investment. This is often adesired intervention outcome. However,earlier and greater investment by theprivate sector might, for example, leadto the same number but lower qualityoutputs and outcomes. In these cases,an explicit adjustment would be neededto weight the higher quality outputsassociated with a public sector onlyoption to ensure that like was beingcompared with like. The weightedoutputs under the various deliveryoptions would then be used to calculatethe cost per unit of adjustedoutput/outcome and be taken intoaccount in the value for moneyassessment.

(iii) valuing the outputs/outcomes – in somecases, outputs are traded and have amarket value. These values can be usedin calculating the additional impact of anintervention. One obvious example iswhere consideration is being given to anintervention, which produces a ratherlow number of net additional jobs, butthese jobs are qualitatively different fromthose that would arise in the referencecase. Often the wage will approximatemany of the quality characteristics of thejobs. Therefore, the level of netadditional earnings (or Gross ValueAdded) can be calculated for eachoption and compared. However, thiswould not be appropriate if the rationalefor an intervention was to createaccessible employment for low skilled,disadvantaged individuals.

A number of examples of how to assessquality additionality are set out in Section5.3.11.

In other cases (for example, many social orenvironmental outputs), valuing a non-marketed commodity is difficult. There are anumber of potential techniques available todo this, such as contingent valuation andhedonic pricing, but these are oftenmethodologically complex and can beexpensive to apply.


Additional impacts

Additional impacts

211

2.7 Maximisingadditionality

Interventions should be designed (or re-designed) so that their additional impact ismaximised. It is unlikely that 100%additionality could be achieved, as not allfactors will be within the control of theintervention developer.

The additionality of an intervention should beconsidered from the earliest stages of aintervention’s development so that wherepossible leakage, displacement andsubstitution are minimised and potentialmultiplier effects are maximised (see Box 2.2across). Wherever possible, interventionsshould be designed to maximise additionally(or ‘design-out’ non-additionality).

Box 2.2: Maximising additionality

Example 1: A project to provide new housing for key workers could minimise leakage ifonly those on an approved list are allowed to occupy the homes.

Example 2: A housing project will reduce the scale of displacement if the type ofaccommodation provided is designed to meet demand that is not being met by the privateor public sector.

Example 3: A project aimed at providing specialist business advice to new start-ups in aspecific area could reduce the amount of leakage by setting out eligibility criteria based onthe project objectives. These might, for example, include target area and size of turnoverand/or number of employees per company.

Example 4: A project aimed at increasing the computer skills of local people couldminimise displacement by checking that there are no other providers of similar trainingcourses either within or outside the target area that the target beneficiaries would be ableto access.

Example 5: A project to assist a large company to locate in a particular area aimed atincreasing employment opportunities for local people could maximise the multiplier effectsby putting in place local procurement and local supply chain development initiatives.

Example 6: A project aimed at providing community facilities in a particular spatial areacould restrict bookings to activities that benefit the target population and thus reduceleakage of benefits.


Additional impacts

Additional impacts

212

3.1 OverviewThe reference case (previouslyreferred to as the base case12) isthe position in terms of targetoutputs and outcomes thatwould occur at the end of theintervention life if the interventionwas not implemented. Thequantification of outputs andoutcomes under the referencecase is referred to asdeadweight.

The reference case is a dynamic concept andinvolves judgement about, amongst otherthings, the economic, social andenvironmental trends or events that areplanned or are thought likely to happen overthe intervention period (but assuming theintervention did not go ahead). The forecastreference case can be better or worse thanthe existing position (known as the baselineposition) depending on the view taken of whateconomic, environmental or social changeswill take place over the intervention period.

It is not an easy task to estimate what wouldhappen in the future if the intervention didnot go ahead and the longer the appraisalperiod the more difficult it is to predict withany degree of accuracy what might happento the target outputs/outcomes comparedto the baseline position.

This section sets out the issues that shouldbe considered when assessing the baselineposition and modelling the future referencecase scenario. It discusses the evidencefrom evaluations and sets out the keyquestion that need to be answered inrelation to deadweight.

3.2 Measuring thebaseline position

An early and key step in carrying out anappraisal is to measure the baseline position

and trends. The baseline is the state of theeconomic, social or environmental context atthe beginning of the intervention period. Thisis usually described quantitatively but,depending on the nature of the intervention,can often also involve qualitativedescriptions of important features. Anassessment of the policy context will alsonormally form part of a baseline assessment.

Interventions designed to address economicand social problems, such as new housing forkey workers in regeneration areas, will need toreview a wide range of quantitative andqualitative baseline and trend indicators,including market and housing needinformation. Interventions with economicobjectives will generally focus on describingthe ‘economic state’ of the target group orarea in terms of, for example, the level ofemployment, unemployment, skills, jobvacancies and industrial classification ofemployers in the travel to work area.Interventions designed to promote communitycapacity building would draw on socio-economic statistics but are also likely todescribe the existing infrastructure and agencyrelationships and other capacity indicatorssuch as attendance at community events.

3.3 Assessing thereference case

The starting position for making anassessment about the likely reference caseis to identify all of the factors that willinfluence the target outputs and outcomes.For different types of outputs/outcomesdifferent factors will be relevant.

Amongst the factors that may need to beconsidered are the following:

Likely changes in social, economic andenvironmental variables;

The nature of the activity beingconsidered;

Evidence from past changes in the localand comparator areas;

The extent of market failure in the areaconcerned;


The reference case (adjusting deadweight)

3 The reference case (adjusting deadweight)13

The referencecase (adjustingdeadweight)

3

12 The Green Book now uses the term base case to mean ‘the best estimate of how much a proposal will cost ineconomic terms, including an allowance for risk and optimism’.

Impacts of health and safety, legal orother statutory requirements; and

Impacts of other relevant existing and/orplanned investments/policies by theprivate or public sector.

Considering the impact of these contextualfactors on the target outputs and outcomeswill enable a reference case to bedeveloped.

For a development project, particularattention should be paid to the assessmentof the reference case when a site has aparticular land-use planning allocation.Where the intervention involves the use ofland, the planning context is likely to be animportant consideration in determining thereference case. Where, for example, theplanning authorities have made itconsistently and unambiguously clear thathousing development will not be permittedthen it is not realistic to suggest that a pieceof land proposed in the intervention optionto be used for commercial development willhave as its reference case a residential useoption. However, a planning allocation initself does not warrant the allocationbecoming the reference case. A marketassessment or other evidence which showsthat the site is likely to be developed forsuch a use is needed. Each case has to bejudged on its own merits.

The best approach to estimating the scaleof target outputs and outcomes under thereference case will normally be to constructa detailed ‘do nothing’ or ‘do minimum’option. This will present what you thinkwould have happened anyway without theintervention, based upon a reasoned casewith supporting evidence. Where it is notpossible for reasons ofproportionality/resources or data limitationsto construct a detailed reference case thenit is possible to use an estimate of theproportion of activity that would haveoccurred anyway.13

It is sometimes seen as convenient andperhaps conceptually easier to assume thatnothing would happen if an intervention didnot go ahead. However, it is highly unlikelythat the current situation (the baselineposition) will remain unchanged over thechosen appraisal period. The variables thataffect local sustainable development arenumerous and constantly changing. Robustjustification would be needed to assumethat the baseline and the reference casewere the same.

When assessing the additional impacts ofan intervention, care has to be taken thatlike is compared with like. Thus, the grossdirect outputs and outcomes generatedunder the reference case must be adjustedfor displacement, leakage and multipliereffects, where relevant, to arrive at a total

net local reference case. The total net localimpacts of the reference case are thendeducted from the total net local effects ofthe intervention options to provide anestimate of their net additional impacts.Linked to this is the treatment of the costsassociated with the reference option. Insome cases this may be zero where there isno public sector intervention. Where thereference case is a do minimum and there isa cost involved, this cost should besubtracted from the intervention cost optionto arrive at a marginal cost. The netadditional outputs should be considered inrelation to the gross and net marginal publicsector cost in the value for moneyassessment.

Table 3.1 sets out a worked example ofestimating the reference case – explainingthe assumptions used to appraise anintervention aimed at providing key workerhousing in a regeneration area over fiveyears. Under the reference case, additionalhousing is expected to be 50 units basedupon historical data and the expectationthat the trend is likely to continue or improveslightly due to housing pressures.Consideration of the remaining additionalityfactors, based on an assessment of thesocial, economic and physical context,reduces this gross direct output from 50housing units to 20 housing units likely to beprovided and occupied by key workers.




314

13 Scottish Enterprise’s ‘Economic Impact Assessment Guidance Note’ sets out a guideline range of values for deadweight, as a percentage of the gross direct effects of the interventionoption, as follows: none – 0%; low – 25%; medium – 50%; high – 75%; and total – 100%. These ready reckoners should only be used where better quality data is unavailable.

Table 3.1: Establishing the reference case for key worker housing

Basis for assumption

Gross direct housing units 50 Local authority data shows that 10 housing units on average have come forward per annum in the target area over the last 10 years

Less estimated leakage -30 The latest Census of Population results show that approximately 40% of those in employment in the area are employed in key services.Previous evidence has been that the proportion of key workers accessing new housing has been broadly similar to this. The project does notplace any restrictions on the characteristics of occupiers.

Gross local direct effects 20

Less displacement 0 Almost all the new housing will be in refurbished existing run down vacant stock. The built up nature of the area means there is littleopportunity for new developments. Displacement is thus expected to be minimal.

Net local direct effects 20

Plus multiplier effects 0 Whilst there may be economic benefits in terms of additional income and jobs arising from the use of local labour and materials in the design, construction and fit out of the new housing this is unlikely to lead to further new housing

Total net local effects 20under the reference case

3.4 Evidence fromevaluations andresearch

Estimates of the level of activity that wouldhave happened anyway are now made as amatter of routine in most evaluations ofregeneration, renewal and regionaldevelopment initiatives. There is thus aconsiderable body of evidence to refer to.The level of deadweight varies considerablyacross programmes reflecting the nature ofthe activity and the local economiccircumstances.

Table 3.2 illustrates that the deadweightassociated with some interventions can behigh – in this case business supportinterventions. In such cases, carefulconsideration should be given to whetherinterventions with high deadweight (and low

additionality more generally) should besupported.

The Final Evaluation of City Challenge (2000)identified a range of estimates of deadweightfor different intervention types, including anallowance for the effect on the timing andscale of activity. The estimates of the level ofdeadweight were based upon two keysources – a beneficiary survey andprogramme and project managerconsultations. Table 3.3 shows thedeadweight estimates by intervention type.The assessment highlights that significantdifferences – for example, in relation tohousing – can occur in the estimates bysource.

Evidence from a recent review ofNeighbourhood Renewal Fund projects,shows relatively low levels of estimateddeadweight (see Table 3.4). This is, inparticular, due to the nature of theprogramme.

3.41 Key question

The key question that needs to beanswered in terms of deadweight is:

What level of outputs and outcomeswould happen anyway without theintervention?

The possible sources of information toanswer this question include:

Evidence from past changes in local andcomparator areas;

Assessments of forecast market,economic and demographic trends;

Local policies and strategies; and

Evidence from previous evaluations andresearch.




315

Table 3.2: Deadweight factors by type of intervention

Intervention type Deadweight factor (%)

Generic Business Support 83%

Access to Finance 65%

Targeted Support 85%

Table 3.3: Estimated deadweight – City Challenge

Intervention type Programme and Project Manager Beneficiary survey-based Overallsurvey-based estimate (average) estimate (average) average

Development schemes 40% 16% 28%

Housing 41% 10% 26%

Transport 37% 12% 24%

Environment and amenity space 39% 21% 30%

Business support 15% 36% 26%

Training and access to labour market 16% 15% 15%

Community and social 23% 15% 19%

Crime prevention 21% 16% 19%

Health 30% 23% 27%

Average across intervention types 31% 17% 24%




316

Table 3.4: Estimate of deadweight – Neighbourhood Renewal Fund

Evaluator’s view

Crime 18%

Education 19%

Health 18%

Housing and environment 24%

Worklessness 20%

Other (including community) 23%

Average 20%

4.1 How to assess theadditionality ofeach option – thefactors explained

The Guide now goes on toconsider leakage, displacement,substitution and multiplier effectsin turn. For each type of effectwe set out:

(i) a simple definition.

(ii) a description of the factors influencing itsscale. As we have already seen the sizeof the target area or area of benefit willsignificantly affect the various factors.

(iii) a review of the various approachesavailable to estimating the scale of eachfactor.

(iv) a brief review of the evidence availablefrom evaluations and other research.

(v) a ready reckoner14 – which simplifies theprocess of assessing the net additionalimpacts by providing a series of estimatesof the scale of each effect. However,project specific information should alwaysbe used in preference to the readyreckoner, where it is available. In addition,evidence should be presented in anappraisal to justify the ready reckonerimpact selected for each effect. The readyreckoner should never be used withoutreference to the project context. Wherethere is uncertainty, it may be helpful touse ranges.

(vi) the key questions to ask as part of aproject appraisal in order to assess eachfactor.

Each of the additionality factors will need tobe applied to both the reference case andintervention option, so that the net additionalimpact can be calculated (see Section 5).

4.2 Leakage4.2.1 Definition

Leakage

The proportion of outputs that benefitthose outside of the intervention’s targetarea or group.

The target beneficiaries for manyregeneration, renewal and regionaldevelopment interventions are individuals,organisations or businesses who form aformal or informal group, based on a sharedcharacteristic or characteristics. Forindividuals these characteristics may, forexample, include key worker status,graduates, ethnic minority, gender and/oremployment status. Frequently interventionsare also designed to benefit groups and/orindividuals living in a particularlocation/community or those in specificindustries or with or without particular skills.The latter may or may not share otherpersonal characteristics. As such, leakage isused to make some allowance fordistributional issues.15

Given the range of regeneration, renewaland regional development type interventionsand the contexts in which they areimplemented, assessing the extent of benefitor output/outcome leakage is often notstraightforward. A number of complex16 andinter-related issues need to be addressed,including:

Users and beneficiaries: there are caseswhere the output/outcome underconsideration may relate to the usage of afacility. In some cases the users andbeneficiaries will be the same – forexample, the users of a communityfacility. In others the target beneficiarymay be indirectly related to users. Thelatter may include the number of touristsvisiting a new facility, where the maintarget beneficiaries are local peoplegaining jobs as a result of visitorexpenditure.



4 Adjusting the reference case and intervention options17

Adjusting thereference caseand interventionoptions

4

14 The ready reckoner draws heavily upon the New Deal for Communities and Single Regeneration Budget ProjectAppraisal and Approval guidance issued by the, then, DETR in October 2000.

15 Not all projects will have solely economic efficiency-type aims. Many will be focused on achieving redistributiveobjectives. Adjusting for leakage will help to ensure that the calculation of net additional impact takes account of theseredistributive concerns

16 There are particularly complex issues relating to the assessment of the other additionality factors. If displacement wasto occur but to a non-target group – for example, from non-Small and Medium sized Enterprises (SME), when thetarget was SMEs – then it could be argued that this displacement should be ignored in relation to this particular

Multiple target beneficiaries: manyinterventions will seek to benefit a rangeof beneficiary groups. For example, a newbusiness incubator may wish toencourage graduates into employmentand also to generate employmentopportunities for disadvantaged localresidents. The leakage rates for thesewould be different.

Leakage of physical outputs: manyappraisers have found it conceptuallydifficult to understand how leakage canrelate to physical asset, such ascommercial floorspace developed. Whereusers are from outside of the target groupthere is logic in reducing the floorspaceclaimed as being additional. However, thishas not normally been done.

Is the area or the individual the target?For example, how far is it the objective ofsmall area regeneration policies toimprove the lot of people who live in thearea and how far to reduce thedeprivation of the area? Thus, anappraiser would need to determinewhether, if a resident secures employmentas a result of the intervention andrelocates, this is leakage or not.

Sources/evidence for estimated leakagefor geographic areas or target groups: thesources/evidence to inform anassessment of the level of leakageassociated with a geographic area or aspecific target group are different. In thecase of the former leakage will usuallyrelate to the place of residence of thebeneficiary – for example, whether theperson gaining a job lives within oroutside of the target area. For jobs thiscan be informed by secondary sourceevidence on travel to work patterns.However, for specific target groups thesources upon which to make evidence-based judgements will often be lessreadily available. As such, in many cases,they will need to be drawn fromintervention specific information (such asthe intervention business plan), analogousinterventions (where data is available), orprimary research.

Leakage implies that no value is attachedto benefits that accrue to non-targetbeneficiaries: where interventions areconcerned with distributional issues thiscan be argued to be logical. However,where the rationale relates to a marketfailure argument and therefore economicefficiency it is not obvious why thesepotential benefits should be discounted,although this may relate as much to howthe beneficiary group is defined. Thepositive and negative impacts on otherareas or groups should also beconsidered in an appraisal.

Leakage in relation to outputs andoutcomes: the leakage of benefits fromtarget groups is likely to be relevant to alloutcomes, but as the above discussiondemonstrates, can be more of an issue inrelation to outputs.

Past experience has been that leakagehas been reasonably consistently appliedin relation to employmentoutputs/outcomes. However, it has eithernot been applied or has been appliedinconsistently in relation to otheroutput/outcome areas. In view of theimportance of targeting particularbeneficiaries in relation to regeneration,renewal and regional developmentinterventions, this Guide recommendsthat leakage be applied consistently to alloutputs/outcomes, including outputssuch as the number of housesdeveloped. As such the precise definitionof the intended beneficiaries is a key partof the additionality assessment andappraisal more generally. Where there isno specific target beneficiary then leakagewill be zero. Thus, for example, if theobjective is to increase take-up of homesin an area and it does not matter who theoccupiers are, then no leakage will occurin this case.

However, as with the other components,the level of analysis and resource devotedto assessing leakage should always berelated to the nature of the investment.Thus, a novel, contentious, repercussive,large and/or complex intervention willrequire more effort, as will one wheredistributed effects are a particularlyimportant objective.

4.2.2 Examples of potential leakageeffects

The potential benefits of an intervention maybe lost to an area or group in a number ofways and the following discussion considersthe ways in which leakage may occur andmay need to be assessed for a variety ofintervention types.

(i) Housing

Interventions designed to provide new orrefurbished housing units will normally needto consider the possibility of leakage. Thekey issue is the relationship between thecharacter of the occupier and the targetgroup. Where the housing units have beenbuilt with the intention of providingresidences for particular groups or peoplefrom a particular area and it is possible thatthese intended beneficiaries will not take upthe accommodation then leakage mightoccur and needs to be assessed.

Another form of leakage that might occurwould be if existing local residents – whowere the target beneficiaries – decided to‘cash in’ and move out of the area.

(ii) Commercial development

This usually involves the reclamation orrefurbishment of existing land or buildings orthe bringing forward of new developmentsto provide increased capacity forcommercial activity.

In terms of the beneficiaries of the building,this may be either the immediate users ofthe building, that is, the companiesoccupying space or those employed by thetenant companies. Where the rationale forthe intervention is to create job opportunitiesfor people in a particular area or targetgroup and not all the space or jobs are likelyto be taken up by those targeted, thenleakage will need to be assessed. Similarly, ifthe development was brought forward withthe intention of providing space for particularindustry sectors or businesses at aparticular stage in their development and theeligibility criteria is such that the possibilityexists that these businesses or sectors donot use all the space then ‘leakage’ mayoccur and needs to be assessed. However,different leakage rates would apply if the




418

target beneficiaries were both local residentsgaining jobs and businesses within aspecific sector.

Where a development takes place with noobjective of attracting a specific group orsector and indeed is keen to attractnewcomers to an area then leakage will bezero.

(iii) Transport

Transport interventions designed to benefitparticular areas or groups of individuals canalso have leakage associated with theoutputs and outcomes they generate. Theimportant point is to be clear about thereasons why the intervention is to beundertaken and what is the target outcome.A new road built to improve access to anindustrial area will not have leakage ofoutputs if the intention was purely toincrease the uptake of development spaceon the site. However, if the primary objectivewas to increase uptake of jobs on the siteby residents in a particular area then there isa likelihood that some leakage of benefitswill occur and these will increase dependingon how accessible the new road makes thesite to non target beneficiaries and whethertheir usage is at the expense of the targetbeneficiaries.

(iv) Business support

An intervention aimed at providing intensivebusiness support to early stage, high-tech,start-ups in the bio-science sector locatedwithin a particular area, is, assuming theeligibility criteria for determining who canreceive support are strictly applied, likely tohave a very small amount of leakageassociated with its outputs and outcomes.However, an intervention providing generalbusiness advice to an unspecified audiencewith the aim of generating jobs in aparticular area is likely to have a greaterdegree of leakage associated with itsactivities as businesses may receive adviceand generate jobs that do not go to targetarea residents or target groups.

(v) Community and social

Interventions aimed at improving the qualityof life of a target group or those living in aparticular area, such as provision of a

community centre, playground or leisurefacility may find it difficult to ‘design out’ allelements of leakage as it may beimpracticable to develop or implement userpolicies that mean that non-targetbeneficiaries are excluded from using thefacility provided. The level of leakage willdepend on the degree to which access canbe controlled. Other communityinterventions such as crèches or healthcentres have the potential to limit usersmore directly by allowing only those within acatchment/target area to register.Nonetheless, there is still the possibility of adegree of leakage as non-targetbeneficiaries may be able to benefit fromliterature/workshops/emergency provisionoffered by the health centre or otheractivities offered by the crèche such as asummer play scheme. Where the healthcentre or crèche serves an area wider thanthe target area, there is likely to be a higherlevel of leakage. Again an importantconsideration will be the extent to which theusage by non-target beneficiaries is actuallyat the expense of use by targetbeneficiaries.

(vi) Training/Education

Training interventions can be developed withthe objectives of improving skills andenabling trainees to gain a qualification. Thiscan be aimed at the population as a whole,or, as is often the case, the training will betargeted at a particular sub-set of thepopulation – such as mothers returning towork, the unemployed, ethnic minorities,graduates, those working in a specificindustry and those in a specifiedoccupation. Training is also frequentlytargeted at those living in a priority area.Even for those interventions with a smalltarget group it should be possible, in theory,to design out leakage with good projectdesign and delivery using appropriateeligibility criteria, rigorously applied. Inpractice, of course, this is likely to provedifficult. Thus, the likelihood of non-targetbeneficiaries taking up training placesshould always be considered and the scaleof potential leakage assessed.

4.2.3 Factors influencing the leakageeffects

The level of leakage will be influenced byfactors such as:

How accessible the intervention outputsare to people from outside of the targetarea or from outside of the target group.This will depend upon both road andpublic transport linkages, as well aspolicies to target usage:

The nature of the output, such as newjobs, that will be created and the ability oflocal residents or a particular target groupto access or to compete for these. In thecase of jobs, for example, this woulddepend upon the skills of the targetpopulation. As an example, if anintervention created local employment inthe retail sector, given the relatively lowskills levels required and low salariesassociated with the sector, it is less likelythat there would be significant interest inavailable positions from outside the localarea. Coupled with this is the likelihood oftheir being a significant pool of suitablepotential employees in the locality.Leakage would therefore be expected tobe low. In contrast, the creation of higherquality jobs is likely to lead to higher levelsof leakage as they provide more incentivefor people from outside the area tocommute in order to access theemployment opportunities; and

The state of the economy in the targetarea – if the intervention is aimed atgenerating economic benefits and theeconomy in the target area is verybuoyant with limited spare resources(labour, capital, etc) able to take up theopportunities offered by the intervention,then leakage may be high as capital andlabour may have to be sourced fromoutside of the target area.

Interventions should be designed to limit thelevel of leakage. Thus, for example,development projects which willaccommodate new employmentopportunities, and where the objective is toincrease local employment, will often needto be combined with a package of trainingsupport for local residents to ensure that




419

they have the skills required by thebusinesses that will occupy the newdevelopments.

4.2.4 Approaches to estimating leakage

In order to estimate the likely level ofleakage, information can be used from thefollowing sources:

Published secondary sources, such astravel to work information;

Local business surveys undertaken by, forexample, the Learning Skills Council(LSC) or local authorities, will sometimesask about the place of residence ofemployees. The local JobCentre Plus canalso be an important source ofinformation upon which to draw;

Labour market studies again produced bythe LSC may also include information onskills and travel to work flows;

Evaluations of previous programmes mayhave included estimates of leakage; and

Surveys/primary research.

4.2.5 Evidence from evaluations andresearch

It is perhaps somewhat surprising that thereis a relatively limited amount of research

relating to the size of leakage effects. Thisundoubtedly reflects the difficult conceptualand measurement problems that exist inseeking to derive good estimates.

Research in the 1980s and 1990s intoproperty driven regeneration initiatives(HMS0 1987, HMSO 1995A and 1995B)revealed that leakage effects dependedheavily on the type of jobs created and thusthe occupations of the people who got thejobs. Thus, the higher the number ofmanagerial, professional and technical staff,the more likely it is that workers fromoutside the area targeted for regenerationwould secure the jobs generated. Mostother occupational groups had around 10%of staff recruited from outside the local areawith the exception of skilled manual workerswhere the equivalent figure is around 20%.The study was also able to ascertain that ingeneral companies in fairly deprived areaswere filling about 40% of their vacanciesfrom unemployed people in the local area.

The Final Evaluation of City Challenge (the,then, DETR, 2000) found that 38% ofemployees in businesses supported by CityChallenge Partnerships lived outside of theCity Challenge area and 11% outside of thelocal authority district.

Relatively low levels of leakage wereidentified through a recent review ofNeighbourhood Renewal Fund projects (seeTable 4.1).

4.2.6 Ready reckoners

Leakage effects can be assessed as inTable 4.2:

If leakage was anticipated to be very high(i.e. 75%) then only 25% of the interventionoutput (i.e. 100% – 75%) would beexpected to benefit members of the targetgroup or those living in the target area ofbenefit.

4.2.7 Key Question – Leakage

In order to address the issue of leakage inan appraisal, the following questions need tobe answered:

Who are the target beneficiaries?

Are the outputs/outcomes likely to benefitnon-target group(s) at the expense of thetarget group(s)? If yes, by how much?




420

Table 4.1: Estimate leakage – Neighbourhood Renewal Fund

Evaluator’s view

Crime 5%

Education 9%

Health 9%


Worklessness 9%


Average 8%

Table 4.2: Leakage

Level Description Leakage

None All of the benefits go to people living in the target area/the target group 0%

Low The majority of benefits will go to people living within the target area/the target group 10%

Medium A reasonably high proportion of the benefits will be retained within the target area/target group 25%

High Many of the benefits will go to people living outside the area of benefit/outside of the target group 50%

Very high A substantial proportion of those benefiting will live outside of the area of benefit/ be non-target group members 75%

Total None of the benefits go to members of the target area/target group 100%

4.3 Displacement 4.3.1 Definition

Displacement

The proportion of interventionoutputs/oZputs/outcomes elsewhere inthe target area.

4.3.2 Examples of displacement

Displacement arises where the interventiontakes market share (called product marketdisplacement) or labour, land or capital(referred to as factor market displacement)from other existing local firms ororganisations. For example, an interventionmay help a business to expand itsoperations. However, this business may takemarket share from other local firmsproducing the same goods or services,resulting in them losing trade and possiblystaff. Alternatively, the supported businessmay use up scarce local factors ofproduction (such as skilled labour) or bid upfactor prices.

In terms of housing, a supported schememay result in a decrease in demand inadjoining areas or elsewhere in the targetarea. Another longer term form ofdisplacement could be the gentrification ofan area, with low income residents beingdisplaced. Displacement may also occur

between tenures – for example, from privaterented to social rented. In the latter case,issues such as the quality ofaccommodation would need to beconsidered in the appraisal.

Another form of displacement may occur ifcrime prevention initiatives cause criminalactivities to happen elsewhere outside of thetarget area.

4.3.3 Factors influencing the scale ofdisplacement

The scale of displacement effects will varydepending upon the nature of activitysupported and local markets. For example,if the supported business has few localcompetitors then the level of product marketdisplacement will be low. In terms of factormarket displacement, an intervention mayresult in an increase in demand forconstruction workers. If these are in shortsupply, the result may be in delays to this orother interventions or an increase in costs.

4.3.4 Approaches to estimatingdisplacement

An assessment of the likely level ofdisplacement can be informed by:

Market analyses: relevant local markets(including product, property and labour)will need to be carefully assessed;

Surveys and studies: some localbusiness surveys will ask questions suchas where are your competitors located

and where are your main markets. Thisinformation can be used to inform anassessment of displacement; and

Evaluations.


There is a considerable body of evidenceconcerning estimates of the scale ofdisplacement associated with regeneration,renewal and regional development initiativesat the local and regional level. The level ofdisplacement at the regional level (NorthEast) associated with various businesssupport activities is set out in Table 4.3. Ahigh level of variation is evident.

There is also evidence that smallercompanies tend to be associated withhigher displacement than larger companies.The reason for this is that small companieswill have more tendency to trade a higherproportion of their output locally than largercompanies (see, for example, the evaluationof TEC Delivered Services, HMSO, 1995).

The Final Evaluation of City Challengeassessed displacement for a number ofintervention types. Displacement wasconsidered to be low at the City Challengelevel, but increased rapidly beyond the localarea (see Table 4.4)17. At the local level,displacement ranged from 8% for trainingand education and business supportprojects to 17% for commercialdevelopment schemes.




421

17 The high levels of displacement at the county, region and UK level reflect the fact that City Challenge was concerned principally with redistribution, rather than removing major supplyside constraints.

17 Displacement/substitution in this case will only apply to jobs created through training as opposed to qualifications gained.

Table 4.3: Displacement – Objective Two Business Support

Jobs Turnover

Generic business support 49% 63%

Access to finance 19% 14%

Targeted support (including new markets, technological development and support for sectors and clusters) 42% 23%




422

The recent review of NeighbourhoodRenewal Fund projects identified similarly lowdisplacement rates to City Challenge at thelocal level (see Table 4.5). In relation to crime,displacement effects principally related to theadverse impacts of the intervention on levelsof crime outside of the target area. Thedisplacement effects in terms of educationand health, on the other hand, were mainlyconcerned with the intervention replacingother public sector provision.


In the absence of specific local informationthe level of displacement can be assessedas in Table 4.6.

If the level of displacement was estimated tobe low (i.e. 25%), then 75% of the outputswould be taken forward (i.e. 100% – 25%).

4.3.7 Displacement and crowding out

There is often confusion betweendisplacement effects and crowding out. Theformer relates to the impact of anintervention on other, normally similar,activities within the target area. The latter isconcerned with macro-economicadjustments that result from an intervention.Crowding out effects are normally onlyconsidered for very large interventions.

4.3.8 Key question – displacement

The following key question needs to beanswered:

Will the intervention/option reduceexisting activity from within (or outside)the target group or area? If yes, by howmuch?

Table 4.4: Displacement rates City Challenge

Intervention type Within City Challenge Immediately adjoining area District County Region UK

Development 17% 21% 38% 71% 89% 91%

Housing 10% 19% 38% 84% 100% 100%

Training and education 8% 17% 31% 77% 78% 80%

Business support 8% 19% 31% 49% 75% 75%

Table 4.5: Displacement rates – Neighbourhood Renewal Fund

Evaluator’s view

Crime 9%

Education 13%

Health 11%


Worklessness 13%


Average 11%

Table 4.6: Displacement

Level Displacement Displacement effect

None No other firms/demand affected 0%

Low There are expected to be some displacement effects, although only to a limited extent 25%

Medium About half of the activity would be displaced 50%

High A high level of displacement is expected to arise 75%

Total All of the activity generated will be displaced 100%

4.4 Substitution4.4.1 Definition

This effect arises where a firm substitutesone activity for a similar one (such asrecruiting a jobless person while anotheremployee loses a job) to take advantageof public sector assistance. It can bethought of as ‘within firm’ displacement.

4.4.2 Examples of substitution

Substitution is a very specific form of non-additionality that has in the past been largelysubsumed within the displacement effectand as a result not considered sufficiently.

If a grant was introduced to encourage localemployers to recruit long-term unemployedpeople, some employers may replaceexisting employees with new workers inorder to secure the grant. This would haveno real impact and such substitution effectsshould be deducted in assessing the netoutput/outcome. However, care needs to betaken when assessing substitution effects ifthe target group are, for example, the long-term unemployed. In this case some degreeof substitution may be consideredacceptable.

Substitution has been an issue for wagesubsidy programmes and work experienceprogrammes. Employers have an incentiveto dismiss unsubsidised workers andreplace them with subsidised workers. Aparticular concern is that the finite duration

of assistance could tempt employers todismiss subsidised workers when subsidiesrun out and bring in a new cohort ofsubsidised workers.

Substitution could be an issue forregeneration programmes if the regenerationstrategy aimed to persuade local employersto recruit more workers locally and fewerfrom outside the area. On the other hand, itmight be argued that non-local workerscould get other work anyway. However, itwould be more of a concern if the attemptto increase local recruitment resulting in thenew local workers taking the place of otherlocal recruits.

Substitution could also arise in relation toother factor inputs such as land andproperty. A firm renting premises could, forexample, take advantage of accommodationprovided by the public sector at a reducedcost by relocating from its current building.In the case of a residential development, adeveloper could switch to undertake apublic sector funded scheme, rather than analternative scheme elsewhere in the localarea.

4.4.3 Factors influencing the scale ofsubstitution

The scale of substitution effects will varydepending upon the nature of the activitysupported, the degree to which substitutionis an intended effect and the ability ofrecipients to engage in substitution where itis an unintended effect. Substitution willtend to be larger, for example, where nocontrols have been established on recipientsregarding the potential substitution activities.

4.4.4 Approaches to estimatingsubstitution

An assessment of the likely level ofdisplacement and substitution can beinformed by:

Direct questioning of recipients – on theirexpected behaviour;

Surveys and studies – of previousinitiatives;

Evaluations – for example, theDepartment for Work and Pensions hascommissioned a number of evaluationsthat have assessed the level ofsubstitution associated with an initiative.A full discussion of concepts and theirapplication can be found in report ESR14, available viawww.dwp.gov.uk/jad/1999/esr14rep.pdf;

Evidence from evaluations and research;


Where there is no specific information onsubstitution the effects shown in Table 4.7could be applied appropriately:

4.4.6 Key question – substitution

The key question in relation to substitution isas follows:

Will the intervention/option result in a firmsubstituting one activity or input for asimilar one to take advantage of publicfunding? If yes, where and by howmuch?




423

Table 4.7: Substitution

Level Substitution Substitution effect

None No substitution takes place 0%

Low There are expected to be some substitution effects, although relatively limited 25%

Medium About half of the activity would be substituted 50%

High A high level of substitution is expected to arise 75%

Total All of the activity would be affected by substitution 100%

4.5 Economicmultiplier effects

4.5.1 Definition

Multiplier effects

Further economic activity (jobs,expenditure or income) associated withadditional local income and local supplierpurchases.

4.5.2 Types of economic multiplier

The economic impact (jobs, expenditure orincome) of an intervention is multipliedbecause of knock-on effects within the localeconomy. Two types of multiplier can beidentified:

A supply linkage multiplier (sometimesreferred to as an indirect multiplier) due topurchases made as a result of theintervention and further purchasesassociated with linked firms along thesupply chain.

An income multiplier (also referred to as aconsumption or induced multiplier)associated with local expenditure as aresult of those who derive incomes fromthe direct and supply linkage impacts ofthe intervention.

A number of impact studies have alsoidentified a longer-term developmentmultiplier associated with the retention ofexpenditure and population in an area.

Many appraisals use a combined orcomposite multiplier. Thus, for example, if atthe regional level the supply linkagemultiplier was 1.1 and the income multiplier1.2, the composite multiplier would be 1.32(i.e. 1.1 x 1.2). Applying the multiplier givesan estimate of the total direct and multipliereffects. For example, say an intervention

created 100 jobs, then the total direct andmultiplier effects would be 132, if thecomposite multiplier were 1.32. Themultiplier effects alone would be 32 (i.e. 100x 0.32).

4.5.3 Factors influencing the scale ofmultiplier effects

The scale of the multiplier effects will beinfluenced in particular by:

Supply linkage multiplier: the extent ofsupply chain linkages in area of analysis.These linkages vary substantially bysector and area;

Income multiplier: the proportion ofadditional income spent within area ofanalysis.

4.5.4 Approaches to estimatingmultiplier effects

There are a number of ways in whichmultipliers can be estimated, including:

Surveys of businesses and employees:businesses can be asked about the localcontent of the purchases they make andthis information can be used to calculatethe local supply linkage multiplier effects,assuming that the proportion ofexpenditure net of non-recoverableindirect taxes incurred on local goods andservices is similar throughout the supplychain19. In addition, estimates can becalculated of the income multiplier usingdata on local consumption patterns in thelocal economy20. Again the assumption isthat behaviour is similar at each point inthe supply chain.

Previous research/evaluations: a numberof previous studies have assessed thescale of multiplier effects. For example,the Scottish Tourism Multiplier Study setsout detailed information on the multipliereffects associated with different types oftourism activity.

Economic models: various commercialand academic organisations havedeveloped models of the nationaleconomy and of local economies. Thesecan be used to assess the scale ofmultiplier effects resulting from a particularinvestment or change in the level ofemployment.

Input-output tables: these tables provideestimates of supply linkages betweensectors and can be used to estimate thesupply linkage or indirect multipliereffects.


The scale of income and supply linkagemultiplier effects vary according to the mixof economic activity that exists in an areaand the type of intervention that is beingundertaken. The Scottish Executive providesinformation on multiplier effects for individualScottish industries, which demonstrates theextent of the difference between varioussectors. For example, the compositemultiplier effect at the Scottish level for theoil process / nuclear fuel industry is 10.5,compared to a composite multiplier of 1.3for other service activities.

Table 4.8 below is based on the extensiveevidence generated by a number of studiesincluding the Evaluation of the EnterpriseZone Experiment. It provides compositeincome and supply linkage multiplierestimates that are appropriate for four typesof property related activity, namely B1 Office,B2/B8 (general industrial/warehousing),Recreation and Retailing. The estimates areprovided for the local area and regional level.At the local level the range is between 1.21and 1.38. At the regional level the range isbetween 1.38 and 1.56. Generally speakingretailing projects generate the lowestcombined income and supply linkageeffects.




424

19 If the purchases made at a particular point in the supply chain is x per annum and a proportion S is spent on local inputs the effects down the remainder of the chain is estimated as:x (1+S+S2+S3……Sn) or x.1/(1-S).

20 If the total net direct and supply linkage multiplier increase in local business turnover is E, a proportion m of this turnover is paid on average in net local incomes, and a proportion qof net local incomes is on average spent on the products of local businesses, then the total impact on turnover, including induced effects, may be estimated as:E(1+mq+m2q2+m3q3….mnqn) or E.1/(1-mq).

For specific sectors and interventions,multiplier values can be higher than thoseshown in the table. For example, The ToyotaImpact Study identified a compositeemployment multiplier at the level ofDerbyshire, Nottinghamshire, Leicestershire,Staffordshire and the West Midlands of 1.6.Recent research by the Dti into broadbandprojects has identified multiplier effectsranging between two to four times the directeffect.


The ready reckoner values in Table 4.9express general ranges at the very local(neighbourhood) level, and the regional level.However, as noted above, the scale ofmultiplier effects can vary substantially.Consequently, careful consideration needsto be given to the appropriate multiplier touse in an appraisal.

4.5.7 Key question – multipliers

The following key question needs to beanswered in relation to multiplier effects:

How many, if any, additional outputs andoutcomes will occur through purchasesalong local supply chains, employeespending rounds and longer term effectsas a result of the intervention/option?




425

Table 4.8: Composite multiplier effect by type of area: site related problems but active private sector

Intervention type Local area Region

B1 Office 1.29 1.44

B2/B8 1.29 1.44

Recreation 1.38 1.56

Retailing 1.21 1.38

Table 4.9: Multiplier effects

Level Multiplier Composite multiplier Composite multiplier(Neighbourhood level) (Regional level)

Low Limited local supply linkages and induced or income effects 1.05 1.3

Medium Average linkages. The majority of interventions will be in this category 1.1 1.5

High Strong local supply linkages and income or induced effects 1.15 1.7




426

5.1 IntroductionThis section sets out how to calculateadditionality. It contains a number ofillustrative worked examples of how toassess additionality for differentintervention types. A worked examplefrom the evidence base is alsopresented. In addition, the ratio of netadditional to gross outputs is alsodiscussed.

The principal focus of the examples in thissection is on calculating, in a quantitativesense, the level of additionality associatedwith an intervention output. Within anappraisal, consideration would also need tobe given to:

Timing effects; and

Quality.

Wherever possible, the additionality ofoutcomes should be a key concern of anappraisal. A qualitative assessment of thelikely level of outcome additionality shouldform part of an appraisal. This would meananswering each of the questions posed inthe preceding section. However, theemphasis in most appraisals is on assessingthe additionality of those outputs that areexpected to lead to the desired outcomes.Applied appropriately the additionalityframework has the potential to significantlyimprove practice. However, failure to do socorrectly could produce partial or misleadinganalyses.

5.2 How to calculateadditionality

In order to calculate net additionality, thelevel of total net local activity under eachoption – intervention and reference case –needs to be assessed. This involves makingadjustments, where appropriate, for leakage,displacement, substitution, and multipliereffects. The total net additional local impactis then calculated by deducting the totalgross additional local effects of the referencecase from the total net local effects of theintervention options.

The calculation of the total net additionallocal impact of an intervention can besummarised using the equation below:



5 Calculating additionality27

Calculatingadditionality

5

AI = [GI x (1-L) x (1-Dp) x (1-S) x M] – [GI*x (1-L*) x (1-Dp*) x (1-S*) x M*]

Where:

AI = Net additional impactGI = Gross impactL = LeakageDp = DisplacementS = SubstitutionM = Multiplier

* denotes reference case and hence deadweight

The net additional impact is therefore theadjusted intervention option minus theadjusted reference case.

The multiplicative formulation described inthe equation represents the relationship inits simplest form. It implies, for example,that leakage effects occur to the sameextent to the gross effects, as they do todisplacement and multiplier effects. Formany interventions, this may be areasonable assumption. However, for others,more specific assessments may need to bemade and detailed, individual calculations ofeach factor prepared.

The calculation should be based uponevidence-based judgements and involvesbeing explicit about assumptions and theexpected implications of a set of actions.

For example, if an intervention was initiatedto create local jobs, the following issueswould need to be assessed:

Leakage – can local people physically getto where the job opportunities areexpected to be? Do local people have thenecessary skills to compete for the jobs?What linked programmes are in place toensure local people can access theopportunities?

Displacement – will the new jobs taken upby local people result in a reduction ofother local people in employment? Havethe potential adverse effects beenminimised by targeting appropriatesectors?

Substitution – will local employers justtake on a local person and releaseanother to take advantage of publicfunding?

Multiplier effects – will those local peoplewho gain employment spend their incomeon goods and services that support localjobs? Will firms purchase more localgoods and services?

A useful aid to calculating additionality is setout within Scottish Enterprise’s EconomicImpact Assessment Guide. Two calculatorsare provided as support tools to assistconsistent calculation.

5.3 Illustrative workedexamples byintervention type

5.3.1 Housing

The public sector is seeking to promote thecreation of new homes, particularly withinmixed-use schemes, in an area containing alarge number of historic buildings. Publicsector support has been requested towardsa intervention involving a former waterfrontmill site that will deliver a mixture ofcommercial development and housing in avariety of new and refurbished buildings.

It is expected that the intervention willdeliver 50 housing units and 2,000 sq m ofcommercial floorspace. Without publicsector support it is likely that one of the oldmill buildings on the site, which is in the beststate of repair, would be brought forward bythe private sector and would deliver 20housing units. No speculative housingdevelopment has taken place on the site,although in the wider area there is evidenceof unsupported private sector developmentalong the rest of the river frontage. This hasbeen reasonably successful to date. It isexpected that the intervention will causesome decrease in the number of housingunits built elsewhere in the target area. Theintervention is targeted on a number ofdeprived communities living in sub-standardaccommodation with associateddisadvantages in the local area. However,the intervention is intended to help to createa more diverse local community byattracting in new residents. Consequently,the leakage of benefits is in this particularcase zero.

Table 5.1 sets out an assessment of the netadditional housing units in sustained or long-term demand generated by this interventionoption at the site level.

Table 5.2 considers the same output at thetarget area level.

As an alternative example, an interventionaimed at improving housing conditionscould fund the refurbishment of existingvacant units in order that they are expectedto be in long-term or sustained demand.Leakage would apply if non-targetcommunity residents occupied therefurbished housing units. On the basis ofpast local evidence, leakage is estimated tobe 20%. If the refurbishment of the 100units means that a local provider who wouldhave built 50 new houses now will build only10 new houses then displacement is some40 housing units. In addition, it is expectedthat 10% of the 100 units refurbished wouldhave taken place anyway through annuallocal authority spend -this representsdeadweight. This intervention would not besufficiently large to result in displacement.On the basis of these assumptions, the totalnet additional local housing units would be32 (see Table 5.3).

Some wider benefits might result as otherresidents in the area carry out improvementsto their properties as a result of theintervention.

Further details of how to assess theadditionality of housing interventions areincluded at Appendix E.

5.3.2 Business support

A business support project is proposed inorder to help create jobs for local people inan area with high unemployment. It isexpected to create 500 full-time equivalentjobs and the intervention’s impact is beingassessed at the neighbourhood level.

Local research by the LSC suggests that,given the recruitment and training supportavailable, local people will take-up most ofthe jobs and thus the level of leakage will below. Without the support (the referencecase) it is estimated that some 80 full-timeequivalent local jobs would be created in thebusinesses supported at the end of theappraisal period.




528

21 See http://www.scottish-enterprise.com/publications/gva_calculator_a.xls and http://www.scottish-enterprise.com/publications/additionality_and_gva_calculator_b.xls




529

22 Housing units are one of a basket of outputs of this project for which the net local additional effects would need to be calculated. The other outputs may include new business start-ups, jobs, business support and environmental outputs.

23 There may be multiplier effects of the expenditure associated with the housing construction, the ‘not applicable’ refers to the likelihood that this spend will result in more housing unitsbeing built.

Table 5.1: Housing development – at site level (units in sustained demand)

Intervention Option Reference Case Additionality

A Gross direct housing units 50 20

B = A*0 Estimated leakage – none 0 0

C = A-B Gross local direct effects 50 20

D = C*0 Displacement – none 0 0

E = C-D Net local direct effects 50 20

F = Not applicable Multiplier N/A N/A

G = E+F Total net local effects 50 20

H = G (Intervention option) Total net additional local effects 30- G (Reference case)

Table 5.2: Housing development – at target area level (units in sustained demand)





D = C*30% Displacement – 30% 15 6




H = G (Intervention option) Total net additional local effects 21- G (Reference case)

Table 5.3: Housing development – at target area level (unites in sustained demand)



B = A*20% Estimated leakage – 20% 20 2


D1 = C*50% Displacement (Intervention option) – 50% 40

D2 = C*0% Displacement (Reference case) – zero 0


F = not applicable Multiplier N/A NA


H = G (Intervention option) Total net additional local effects 32- G (reference case)

There are a number of other competingfirms in the area and the level ofdisplacement is therefore expected to bemedium. The businesses concerned areknown to have strong local supply linkagesand thus the multiplier effects areanticipated to be high. Leakage,displacement and multiplier impacts similarto those under the intervention option wouldbe expected to apply to the reference caseposition.

Using the ready-reckoners set out in Section4 the estimated level of total local netadditional jobs can be calculated as shownin Table 5.4.

Therefore the business support project isanticipated to create some 218 total netadditional local jobs.

In undertaking an appraisal of a businesssupport project, consideration would alsooften need to be given to the net additionaloutcomes generated, such as increasedeconomic activity (Gross Value Added).

5.3.3 Commercial development

As part of a programme to tackle under-useand dereliction within a run-down CityCentre, public sector support totalling£800,000 is being considered towards thedemolition of a derelict building andconstruction of 8,000 sq m of workspace.The objective of the scheme is to bring neweconomic activity and jobs for local peopleinto the City Centre and adjoining wards.The area has a large number of under-usedand empty, derelict buildings. Over the last10 years little commercial development hastaken place either of a pre-let or speculativenature. The workspace will comprise offices(3,000 sq m) and industrial floorspace(5,000 sq m).

The local authority and Business Link hasadvised that there is evidence of unmetdemand from businesses in the area. Theyhave also indicated that the businessesoccupying the new space are likely to be inthe lower end of the skilled manufacturingand service sectors, although some hi-techbusinesses might locate there. Displacementis expected to be low, albeit slightly higher interms of floorspace than jobs. Without the

intervention it is estimated that 1,500 sq mof floorspace would be refurbished andbrought back into industrial use anyway.

Due to the historically low economic activityand the multiple social problems, the areahas become the target for a number ofpolicy initiatives, with a total investment of£10 million. The building and site are readilyaccessible by public transport and withinwalking distance of a number of residentialareas in the local travel to work area, someof which are within the top 20% mostdeprived in the country. It is also accessibleby car and public transport to otherresidential areas outside of the local area.

Table 5.5 sets out the estimate of the netadditional floorspace created.

The calculation of the number of netadditional jobs created is summarised inTable 5.6.

The outcomes associated with commercialdevelopment will include net additional localemployment and GVA generated by theintervention.

Alternatively, if the intervention option willresult in the same number of employmentopportunities as the reference case, but of ahigher standard, then it may be appropriateto measure additionality in terms of totalincome from employment in the local area.However, this will depend upon theobjectives of the intervention. For example,this approach may well not be appropriatefor a regeneration intervention that isseeking to create accessible jobs, whichmay be low paid.

The public sector is considering aninvestment in the development of 1,000 sq m of office space. Without public sectorintervention, a private sector developerwould construct a warehousing building ofthe same size on the site. In this example,we have assumed that if the building isoccupied for warehousing use, then usingEnglish Partnerships’ employment densities,it will accommodate 20 jobs, many of themwithin lower order occupations. As officeaccommodation, it will accommodate 52jobs with a high proportion of business andpublic service professionals. Leakage is

assumed to be higher under the interventionoption as the skills requirements are lesslikely to be met at a local level than underthe reference case. Income generatedthrough multiplier effects is calculated usingthe regional average annual earnings figure.In this example, the total additional localincome accruing to local employees is£210,000 per annum.

In examples such as this, it will be for theappraiser to determine which occupationalgroups to use to calculate income arisingthrough the development, or whether acombination of occupational groups shouldbe used.

An alternative approach would be to usefigures for turnover per head or GVA perhead by sector, which can be derivedthrough such sources as the AnnualBusiness Inquiry.

5.3.4 Transport

Public sector regeneration funding is beingconsidered towards the construction of anew road, which will link a new socialhousing development and an existingresidential area with an established industrialarea and the main public transportinterchange, which is about to be extended.The objective is to enable current isolatedcommunities and disadvantaged individualsto access employment and otheropportunities. The road will also enablepedestrians and cyclists to access theindustrial areas more quickly and moresafely than the existing route, which wouldrequire crossing a busy dual carriageway.

The road will be approximately 1.5km longand will become an adopted road oncompletion. There are no other public sectoror private sector funders. The targetbeneficiaries are local residents who areexpected to account for 80% of usage –giving a leakage rate of 20%.

Given the main users of the road andpurpose of the intervention, it is notexpected that the road would beconstructed by the private sector in theplanned location or that any alternative routewould be constructed in the foreseeablefuture that would serve the same purpose.




530




531

24 The number of jobs able to be accommodated calculated using standard floorspace to employment density ratios, and allowing for an 80% occupancy rate.25 The number of jobs able to be accommodated calculated using standard floorspace to employment density ratios.

Table 5.4: Business support – employment


A Gross direct jobs 500 80

B = A*10% Estimated leakage –10% 50 8




F = E*(1.15-1) Multiplier – 1.15 34 5


H = G (Intervention option) -G (Reference case) Total net additional local effects 218

Table 5.5: Commercial development – floorspace (sq m)


A Gross direct floorspace (sq m) 8,000 1,500

B = A*25% Estimated leakage -25% 2,000 375

C = A-B Gross local direct effects 6,000 1,125

D = C*20% Displacement – 20% 1,200 225

E = C-D Net local direct effects 4,800 900

F = not applicable Multiplier N/A N/A

G = E+F Total Net local effects 4,800 900

H = G (Intervention option) - G (Reference case) Total net additional local effects 3,900 sq m

Table 5.6: Commercial development – employment


A Gross direct jobs 230 ? 97 ?

B = A*25% Estimated leakage – 25% 58 24




F = E*(1.1 -1) Multiplier – 1.1 15 6


H = G (Intervention option) -G (Reference case) Total net additional local effects 93

Table 5.7: Commercial development – income


A Gross direct jobs 52 20

B1 = A*65% Estimated leakage (intervention option high) – 65% 34

B2 = 10% Estimated leakage, reference case – low) – 10% 2




F = E*(1.1-1) Multiplier – 1.1 2 2


H1 (Intervention option) Average annual earnings for office use (£) 33,000

H2 (Reference case) Average annual earnings for warehousing use (£) 19,000

I Average annual earnings in the region (£) 25,000 25,000

J = (ExH) + (FxI) Total net local effects 545,000 335,000

K = I (Intervention option) - I (Reference case) Total net additional local effects £210,000

It is possible that the owner of the industrialarea may construct a small portion of theroad to further facilitate road traffic accessinto and out of the industrial park. However,under the reference case only a limitednumber of the users would be from the localcommunity (a leakage rate of 80%). Theadditionality of the intervention outputs, interms of number of trips by targetbeneficiaries is assessed in Table 5.8.

The net additional local change inaccessibility and the associated benefits(such as enhanced quality of life andincreased economic activity) would need tobe assessed in order to measure outcomeadditionality.

5.3.5 Tourism

The creation of a new museum is proposed,on a key site within a major city. It isintended that the intervention will stimulateregeneration within the surrounding area bycreating a significant additional culturalattraction and tourist draw. Overall, it isexpected that the intervention would create300 jobs. If the intervention were not to goahead it is envisaged that the existingdevelopment would remain on the site forthe foreseeable future. As such, under thereference case some 100 jobs would besafeguarded.

An analysis of the anticipated level of jobsbenefiting residents within the sub-region(target area) suggests that leakage underthe proposed intervention will be low, with aleakage rate of 10%. Whilst, it is likely thatthe new Museum will draw some visitorsaway from existing attractions, it is intendedthat the Museum will represent a uniquetourism product that is not offeredelsewhere in the sub-region. As such, theoverall level of displacement is expected tobe low, at 25%. A medium to high levelmultiplier effect, of 1.5, is considered to beappropriate for the proposed intervention.This has been based upon a review ofevidence relating to the extent of multipliereffects within creative industries (see, forexample, DCMS’s Creative IndustriesCreative Bulletin, 2004).

Leakage under the reference case is 5 jobs,based upon interviews with existingemployers. The continuation of the existingactivity is not assumed to result indisplacement effects. A multiplier of 1.3 hasbeen estimated, again as a result ofinterviews. Table 5.9 summarises the netadditional number of jobs created orsafeguarded under the proposedintervention after taking account of leakage,displacement, multiplier effects anddeadweight.

5.3.6 Environmental

A request for public sector support hasbeen received to reclaim a two hectare sitecurrently used for unofficial fly-tipping inorder for it to be brought back into use asan adventure playground and country park.There is substantial support for this at thelocal level as there are few alternativefacilities nearby. Over the years the site hasbecome increasingly neglected andunsightly, it is also a health hazard. Theproject site is adjacent to a number ofresidential areas and is easily accessible bypedestrians, cyclists and motorists. Theareas from which pedestrians and cyclistsare likely to come are high priority areas interms of social need. Car borne users mightcome from a wide range of areas. However,the facility is not targeted at any specificgroup or area. Under the reference case,without public sector regeneration support,the site would be fenced off in order toprevent further fly-tipping, although a smallplayground would be provided.

The additionality of the hectares of landreclaimed for soft end use is assessed inTable 5.10.

The outcome additionality associated withenvironmental interventions would need toconsider, for example, the net additionallocal impact on quality of life. Alternatively,measures such as the enhancement inproperty values or willingness to invest couldbe considered for the intervention andreference case options.

5.3.7 Community and Social

A comprehensive package has beendeveloped aimed at addressing the social,environmental and economic issues facedby a rural area. As part of the package, thepublic sector is appraising an interventioncomprising the acquisition of a site,construction and operation of a 250 sq mcommunity centre. The Centre will be usedfor a wide range of purposes, includingtoddler groups, after school club, training inliteracy and numeracy, outreach for Citizen’sAdvice Bureau (CAB) sessions, communitymeetings, lectures and events and it will alsoenable computer training and act as aninformation access point. Many of theseactivities, such as CAB sessions, do notcurrently take place in the local area.

Table 5.11 shows the calculation of theadditionality of the number of communitymeetings involving more than 5% of targetpopulation26. It is estimated that fivemeetings of this scale would take place inthe existing parish hall. In addition, five ofthe meetings held in the new centre wouldbe primarily for non-local residents.

The capital cost of the centre will be fullyfunded by the public sector and incomefrom charges will help to pay some of therunning costs. The target area, which thecentre is intended to serve, is quite largegeographically, containing five small villageswithin a five mile radius of the centre with acombined population of 1,500 people. It isexpected that the majority of users willcome from the target area although it ispossible that for a small number of theevents and activities a number of the userswill come from outside of the target area.For example, the CAB sessions mightattract users from outside of the target area,as might a number of training sessions. Inthe absence of the centre the small parishhall would continue to be used as it is atpresent for a limited range of local activitiessuch as the toddler group and lectures oflocal interest. It is likely that the parish hallwill continue to be used at the same levelregardless of whether or not the centre is built.




532

26 This is one of a number of relevant outputs where the additionality of the project’s outputs would need to be assessed. The other might include numbers of trainees and crècheplaces provided/used.




533

Table 5.8: Additionality of trips – annual number of trips by target beneficiaries


A Gross direct trips on new road 50,000 20,000

B1 = A*20% Estimated leakage (intervention option) – 20% 10,000 N/A

B2 = Ax80% Estimated leakage (Reference case) – 80% N/A 16,000


D = Not applicable Displacement N/A N/A

E = C-D Net local direct effects 40,000 4,000


G = E+F Total net local effects 40,000 4,000

H = G (Intervention option) -G (Reference case) Total net additional local effects 36,000

Table 5.9: Employment additionality arising from the redevelopment of a museum


A Gross jobs 300 100

B = A*10% Estimated leakage (intervention option) -10% 30 N/A

B = A*5% Estimated leakage (reference case) – 5% N/A 5


D1 = C*25% Displacement (intervention option) – 25% 68 N/A

D = 2C*0% Displacement (reference case) – 0% N/A 0


F = E*(1.5-1) Multiplier (intervention option) – 1.5 101 N/A

F = E*(1.3-1) Multiplier (reference case) – 1.3 N/A 29


H = G (Intervention option) - G (Reference case) Total net additional local effects 179

Table 5.10: Additionality of land reclaimed


A Gross direct hectares of land reclaimed 2 0.25


C = A-B Gross local direct effects 2 0.25


E = C-D Net local direct effects 2 0.25


G = E+F Net local effects 2 0.25

H = G (Intervention option) - G (Reference case) Total net additional local effects 1.75

Table 5.11: Additionality of community meetings


A Gross direct number of meetings total 50 5attendance of more than 5% of target population

B1 = A*10% Leakage (intervention option) – low (10%) 5 0

B2 = A*0 Leakage (reference case) – none





G = E+F Net local effects 45 5


The additionality of the number ofcommunity users of CAB outreach servicesin shown in Table 5.12. It is estimated that66 individuals would use the CAB servicesat the new centre, compared with 11 underthe reference case.

Again, the outcome associated with suchinterventions would principally be basedaround net additional improvements toquality of life. For very large interventions, itwould be possible to use survey-basedcontingent valuation exercises to measuresuch effects. However, the results of suchsurveys are likely to provide an overestimateof the overall effects. For example, if localresidents were asked how many times permonth they would expect to use a proposednew community swimming pool they mayoverstate expected usage in order to ensurethat the development proceeds.

5.3.8 Crime prevention and communitysafety

A potential investment in Closed CircuitTelevision (CCTV) is being considered tohelp reduce the level of car theft, burglaries,violent attacks and street crime. The policehave been consulted and they have advisedthat the measure is likely to be effective andcould, for example, reduce the number ofcar thefts by 50% from 100 per annum to50 or fewer. No other source of funding isavailable. The Police have advised that theywill shortly be implementing a number ofnew initiatives that are aimed at reducingcrime in the area, with a target reduction ofat least 10%. If the CCTV were installedthose new initiatives would not happen. Thetarget beneficiaries are those suffering fromcrime. Crime could be displaced to otherareas. However, in this case, suchdisplacement is not expected to happen.

Table 5.13 sets out a calculation of theadditionality of the reduction in car thefts27.

As a further example, an intervention aimedat reducing fear of crime might install locksand entry phones in 50% of the 200 housesin the target community. If 10% of homesinstalled phones and locks at their own

expense then deadweight is 10% (20homes). Leakage is likely to be zero as theeligibility criteria will limit installation tohomes in the target area. There are alsowider effects that might be relevant andworth considering where an intervention’ssuccess encourages those outside thetarget area to adopt new practices. In thisinstance if the adjacent communityrecognised the benefits of installing locksand entry phones and 10 locks and phoneswere fitted, which would otherwise not havebeen, then this would increase the additionalimpact of the intervention, if the target areaalso included these homes. However, theinitiative may have the effect of displacingcrime to other areas and thus potentiallyincreasing the fear of crime in these areas.

5.3.9 Training and education

An information technology trainingprogramme is proposed, targeting a specificneighbourhood. Evidence of residence in thetarget area will be a criterion for eligibility toensure no leakage of benefits outside of thearea. This course will involve the provision ofone week (30 hours) of intensive training pertrainee in a range of software packages,together with job search support. There arealready a number of training providersserving the local area, although the nature ofthe training is more limited in its scope andduration and it is not expected that thisintervention will cause a reduction indemand for the existing training provision. Inappraising the intervention it will also beessential to consider these qualitativeaspects of the intervention. It is estimatedthat of the 40 trainees, perhaps eight ofthem would have undertaken anothercomparable course available elsewhere inthe absence of this intervention.

Table 5.14 shows the calculation of theadditionality of the number of trainees.

As a further example, an intervention aimedat improving school attainment levels mayimprove school facilities. Where it isexpected that a portion of these facilitieswould have been improved without the

intervention going ahead this would bedeadweight. If as a result of the newfacilities, existing facilities weredecommissioned before they were no longerfit for purpose then this would bedisplacement. If the facilities were used bynon-school users then this could be leakagedepending on the objectives and targetbeneficiaries. The size of the multiplier effectwould depend on the amount of local labourand local materials used in the constructionand operation of the new facilities.

5.3.10 Health

An intervention aimed at lowering mortalityrates may build a local community healthcentre. A full appraisal of such anintervention would involve consideration ofhealth impacts. This would normally takeaccount of changes in life expectancy(including expected life years where lives arelost or saved) and changes in quality of life.This approach is known as the quality-adjusted life year (QALY). However, in thecase of this example, we are consideringonly the additionality of the usage of thefacility.

If existing facilities had to close downbecause of the new facilities there would besome displacement. It would also benecessary to consider how the targetpopulation might otherwise have obtainedmedical advice. All non-target users wouldbe classified as leakage. The wider effectsmight be that a non-target but prioritypopulation in an adjacent area adopt thegood health practices of the targetcommunity as a result of the health centre.




534

27 Car theft is again just one of a number of relevant outputs that could measure the additionality of the project, others include, reduction in burglary and personal attacks and in the fearof crime.




535

Table 5.12: Additionality of CAB usage


A Gross direct 66 11

B = estimated leakage Leakage (specific estimate) 13 0







Table 5.13: Additionality of reduction in car theft


A Gross reduction in car thefts 50 10






G = E+F Net local effects 50 10


Table 5.14: Additionality of training places


A Gross direct trainees 40 8



D = Not applicable Displacement 0 0





Table 5.15: Housing development – minimum threshold



B1 = A*0% Below minimum threshold (intervention option – none) 0 -

B2 = A*50% Below minimum threshold (reference case – 50%) - 75

C = A-B Gross direct housing units 150 75

D = C*0 Estimated leakage – none 0 0

E = C-D Gross local direct effects 150 75

F = E*25% Displacement 38 19

G = E-F Net local direct effects 112 56

H = Not applicable Multiplier N/A N/A

I = G+H Total net local effects 112 56

J = I (Intervention option) - I (Reference case) Total net additional local effects 56




536

28 It should be noted that these figures are based upon current available information. English Partnerships’ and others are reviewing this work and as such the estimates of levels ofCO2 may be subject to revision.

29 Source: DEFRA (2007)

5.3.11 Quality

Minimum thresholds

Public sector funding is required to supportthe development of a residential scheme ata site on the edge of the city centre. It isintended that, in total, some 150 units willbe delivered, each of which will be to a highstandard of design and achieve asustainability rating of Code Level 3.

In the absence of public sector support, it isstill thought likely that the site would bebrought forward for residential use and thatthe number units created would be thesame as under the intervention option.However, due to the cost implicationsassociated with achieving a sustainabilityrating of Code Level 3, it is expected thatunder the reference case only 50% of theresidential units delivered would meet thisstandard.

Table 5.15 sets out the estimated number ofnet additional housing units associated withthe intervention option, after considerationhas been given to the quality of the outputscreated.

The extent of public sector support requiredto secure outputs that meet or exceed theminimum threshold should be comparedagainst unit cost benchmarks, in order testfor value for money. For example, a housingscheme delivering homes at Code Level 4might have a public sector unit cost of£35,000 per Level 4 home, which can bebenchmarked against other comparableschemes.

Weighting the outputs/outcomes througha scoring assessment

The development of new public realm isproposed in the form of improvements to atown centre’s main high street. Theintervention will principally concern works toenhance the local environment and includenew, distinct, high quality open space andlandscaping, as well as the creation ofpedestrian areas and public art. Overall,some 3,000 sq m of public realm will bedeveloped.

Under the reference case, it is envisagedthat improvement works to the high streetwould be undertaken, but to a lowerspecification of design. Less priority wouldbe given to pedestrian use and theincorporation of public art, and theemphasis on creating a distinct ‘sense ofplace’ would be lost. The quantum of publicrealm developed would still be expected tobe approximately 3,000 sq m, although thequality of this space would be poorcompared to the intervention option.

The objective of the intervention is to createan environment that will attract additionaleconomic activity to the town. The quality ofpublic realm is therefore important.Consequently, the outputs under eachoption have been scored in terms of theirimpact on the image of the town, basedupon the following scoring range:

9-10 = an extremely significant positiveimpact;

7-8 = a significant positive impact;

4-6 = a positive impact;

1-3 = a marginal positive impact; and

0 = a neutral/no change position.

The additional output score generated undereach option are shown in Table 5.16.

Where more than one output is beingconsidered, it may be appropriate to weighteach output according to its relativeimportance. One approach to presenting aweighting and scoring analysis of multipleoutputs is in the form of a summary spiderdiagram, as shown below.

Valuing the outputs/outcomes

A residential led scheme is proposed thatwill create 100 new housing units on aformer derelict site, within the town centre.As part of meeting the requiredenvironmental standards (Code Level 3), theenergy use of each dwelling will beminimised through improvements to thebuildings fabric to reduce energy demand,along with the provision of efficient energy

supply and renewable energy sources.

If no public sector support is provided, thesame number of units would be constructedas under the intervention option, but theywill be built to a lower environmentalstandard. Consequently, the energyconsumption associated with these units willbe greater, leading to higher levels of carbondioxide (CO2) emissions. It has beenassumed that, on average, the residentialunits delivered under the reference case willemit 1.8 tonnes of CO2 per annum. Incomparison, it is expected that theintervention option will achieve a 25%reduction in energy consumption (e.g. 1.35tonnes of CO2, and hence carbonemissions, per dwelling.

The value of reducing CO2 emissions hasbeen applied to each option, based upona shadow price per tonne of carbonemitted of £26, to provide a total socialcost saving per annum. Table 5.17 setsout the results of this analysis.

*It has been assumed that the unitsdisplaced under the intervention optionwould have been of a similarenvironmental standard and thereforewould have resulted in the same level ofsocial cost saving per unit. If the outputsdisplaced are of a lower quality, thisshould be reflected by a reduceddisplacement rate.

109876543210

109876543210

Economic performance of local area

Image / external perception

Sense of identity and place

Place vitality

Social inclusionSocial interaction

Community safety / crime

Health

Access to goods & services

Heritage value

Energy efficiency

Waste management / pollution

Greenery / ecologyRent / capital value

Market attractiveness

Whole life costs

User performance




537

Key

Red line denotes intervention option

Blue line denotes reference case

Table 5.16: Public realm development – Output score


A Gross direct sq m 3,000 3,000

B = A*0% Estimated leakage – none 0 0


D = C*0% Displacement – none 0 0

E = C-D Net local direct effects 3,000 3,000


G = E+F Total net local effects 3,000 3,000

H1 = Output score (intervention option) 7 -

H2 = Output score (reference case) - 3

I = G*H Weighted total net local effects 21,000 9,000

J = I (Intervention option) - I (Reference case) Total net additional local effects 12,000

Table 5.17: CO2 emissions – valuing the outputs





D = C*25% Displacement* 25 25




I1 = G*1.35 CO2 emissions p.a. (intervention option) 101 -

I2 = G*1.8 CO2 emissions p.a. (reference case) - 135

= I*£26 Social cost p.a. £2,626 £3,510

J = I (Intervention option) - I (Reference case) Total net additional local effects (annual savings p.a.) £884

Summary Scoring Chart

5.4 A workedexample from theevidence base

The estimated level of gross to netadditional turnover associated with variousbusiness support activities funded throughObjective Two in the North East are shownin Table 5.18.

5.5 Gross to netadditionality ratios

Another way of comparing the additionalityof interventions is to consider their gross tonet additionality ratios – the net additionaloutputs as a percentage of gross outputs.Under this approach, interventions whichdemonstrate high ratios would be morebeneficial in terms of additional outputs thaninterventions with lower gross to netadditionality ratios, where the gross directeffects are the same.

The use of gross to net additionality ratios isan alternative approach to calculatingadditionality, which is simpler than derivingindividual estimates for deadweight, leakage,displacement and multiplier effects.However, this approach should only be usedwhere a rough estimate of additionality isrequired – for example, at the interventiondevelopment stage. A detailed appraisalshould include a full assessment of each ofthe additionality factor.

A number of evaluations have calculatedgross to net additionality ratios. Forexample, the recent Interim Evaluation of theCoalfields Regeneration Programmes inEngland (SQW, 2007) identified an overalladditionality rate of 70% – 80%.

The mid-term report into 10 SingleRegeneration Budget case studies30

included details of aggregated additionalityratios by intervention outputs. These rangedfrom 30% for the number of full-timeequivalent jobs safeguarded to 80% for thenumber of employee volunteering schemes.Net additionality ratios for 61 differentoutputs are set out in Table 5.19.

Table 5.20 shows the calculation of grossdirect to total net additional local impacts fora range of Neighbourhood Renewal Fundproject types.




538

30 Department of Transport, Local Government and the Regions (DTLR) – Neighbourhood Regeneration: Lessons and evaluation evidence from ten single Regeneration Budget casestudies. (Number 1, 2002)

Table 5.18: Business support – adjusting from gross to net additional impacts – turnover

Generic business support Access to finance Targeted support

Gross £57.4m £82.4m £84.9m

Less deadweight (less 80%) (less 65%) (less 85%)

Net £11.5m £28.6m £12.0m

Less displacement (less 63%) (less 14%) (less 23%)

Net additional £4.3m £24.7m £9.3m

% of gross 8% 30% 11%




539

Table 5.19: Summary of net additional outputs for the ten case studies (all years)

Description Net additional outputs as % gross

Jobs, training and education

1a1 No FTE jobs created 32

1a2 No FTE jobs safeguarded 30

1a3 No FTE construction jobs (person weeks) 33

1b No pupils benefiting from projects assigned to enhance/improve attainment 54

1c No people trained obtaining qualifications 58

1d No residents accessing employment through training advice or targeted assistance 55

1e No training weeks 62

1f1 No people trained obtaining jobs 48

1f2 • Who were formerly unemployed 45

1g No entering self-employment 46

1j No young people benefiting from projects to promote personal and social development 48

1k1 No employers in collaborative projects with educational institutions to improve student performance 62

1k2 No students in collaborative projects 50

1l No teachers who have had a placement into business in the last period 60

Economic growth

2a No new business start-ups 31

2b1 Business/commercial floorspace improved (m2) 27

2b2 New business/commercial floorspace (m2) 44

2c1 New businesses supported 36

2c2 • Surviving 52 weeks 36

2c3 • Surviving 78 weeks 35

2d No businesses advised as a result of SRB assisted activities 33

Housing

3a1 No private dwellings completed 48

3a2 No private dwellings improved 45

3a3 No local authority dwellings completed -

3a4 No local authority dwellings improved 37

3a5 No housing association dwellings completed 39

3a6 No housing association dwellings improved -

3b No dwellings in tenant management organisation 65

Community safety/crime prevention

5a1 No benefiting community safety initiatives 53

5a2 • Aged over 60 46

5a3 • Females 44

5b1 No dwellings with upgraded security 51

5b2 No commercial buildings with upgraded security 56

5c No community safety initiatives 54

5d1 No youth crime prevention initiatives 51

5d2 • Nos attending crime prevention initiatives 63




540

Table 5.19: Summary of net additional outputs for the ten case studies (all years) (continued)

Description Net additional outputs as % gross

Environmental improvements

6a Land improved / reclaimed for open space (ha) 58

6b Land improved / reclaimed for development (ha) 58

6c No buildings back into use 53

6d1 Roads built (km) -

6d2 Roads improved (km) 67

6e No traffic calming schemes 60

6f No waste management / recycling schemes 50

Community facilities

7a1 People access to new health facilities 66

7a2 People with access to new sport facilities 66

7a3 People with access to new cultural facilities 66

7a4 No new health facilities 60

7a5 No new sports facilities 64

7a6 No new cultural facilities 65

7b1 No using improved health facilities 66

7b2 No using improved sports facilities 62

7b3 No using improved cult facilities 66

7b4 No health facilities improved 52

7b5 No sports facilities improved 61

7b6 No cultural facilities improved 63

Voluntary / community sector

8a1 No voluntary organisations supported 68

8a2 No community organisations supported 75

8c No individuals involved in voluntary work 65

8d No employee voluntary schemes 80

8e No community enterprise start ups 77

Childcare

10a No childcare places provided 65

Table 5.20: Gross to net additionality ratio – Neighbourhood Renewal Fund (Evaluators views)

Gross direct to local net additional local %

Crime 71%

Education 64%

Health 66%


Worklessness 63%

Other 62%

Average 66%

This Guide has explained how to appraisethe additional impacts of an intervention,using evidence-based judgements.

The preceding discussion has made it clearthat the assessment of additionality forms acritical part of any appraisal. It is recognisedthat assessing additionality is not always astraightforward process, and it requiresknowledge of the intervention andjudgement as well as information on whichto base an assessment of leakage,displacement, substitution, multiplier effectsand deadweight. However, without anassessment of additionality we do not knowwhat the intervention is adding over andabove what would have happened anyway.In the absence of this information we cannottell if the intervention offers good value formoney.

This document has identified a number ofissues that must be addressed if anintervention’s additionality is to be correctlyassessed and where possible maximised:

(i) What would happen anyway?(The reference case);

(ii) Who do we want to benefit and willthey?

(iii) Will activity elsewhere in the target groupor area be reduced?

(iv) Will there be additional benefits as aresult of further expenditure?

As well as scale effects (i.e. the quantity ofoutputs/outcomes), it is important to alsoassess whether an intervention will result ina different quality of and/or timing ofbenefits.

Good practice is to always use valuesderived from local experience and research.In the absence of such primary information,intervention developers and appraisers mayon occasions need to use ready reckonervalues referred to in this Guide for thedifferent additionality factors. Where theseare used justification will be needed as totheir appropriateness. They must not beused as replacements for detailed,intervention specific knowledge andresearch. In assessing additionality, theimportant thing is not to calculate a

spuriously precise figure, but rather to beclear about the likely scale and nature of anintervention’s additional impacts. Like manyother aspects of economic appraisal it ispossible to contrive figures – using theadditionality assessment in this way is awaste of time. Where there is uncertainty, itmay be helpful to consider using ranges.

The process of assessing additionality ismore than an input into the value for moneyjudgement. It is relevant to all stages of aninterventions’s lifecycle. It can be used in apositive way as a tool that the intervationdeveloper should use to test the interventionas it is developed, with a view to re-designing it so that:

As many of the outputs as possible reachthe intended beneficiaries;

Existing outputs/outcomes are notunintentionally displaced;

Linkages are made to maximise thebenefits; and

The public sector does not supportactivity that individuals or the privatesector or other organisation would havedone anyway.

Overall, the assessment of additionality is animportant element in maximising the impactand value for money of an intervention andensuring that it delivers real results.


Conclusion

6 Conclusion41

Conclusion6


Conclusion

Conclusion

642

Additionality The extent to which an activity is undertaken on a larger scale, takes placeat all, or earlier, or within a given geographical area as a result of theintervention. Thus, an impact arising from an intervention is additional if itwould not have occurred in the absence of the intervention.

Deadweight Output that would have occurred without the intervention.

Displacement The proportion of intervention outputs accounted for by reduced outputselsewhere in the target area.

Intervention Project, programme or policy implemented or supported by the publicsector to achieve its objectives.

Leakage The proportion of outputs that benefit those outside of the intervention’starget area or group.

Market failure A situation where barriers exist to the normal and efficient operation of alocal economy. Examples may include information barriers, where localpeople do not know about nearby job vacancies.

Multiplier effects Further economic activity (jobs, expenditure or income) associated withadditional local income, local supplier purchases and longer term effects.

Outcomes The wider effects or impact on an area of an intervention, for example thereduction in crime level over a set period of time.

Outputs The physical products or measurable results of individual projects, forexample, the number of firms assisted and training places taken up.

Reference case The position in terms of target outputs over a set period of time if theintervention did not take place.

Substitution Where a firm substitutes one activity for a similar activity (such as recruitinga different job applicant) to take advantage of public sector assistance.

Target area The area within which benefits will be assessed.


Glossary

Glossary43

Glossary

AMION (2007) National Strategy for Neighbourhood Renewal – Evidence onProgress and Value for Money

CLG (2007) Adjusting for Optimism Bias in Regeneration Projects andProgrammes: A Guidance Note

DETR (2000) Final Evaluation of City Challenge

DETR (October 2000) New Deal for Communities and the Single RegenerationBudget Project Appraisal and Approval

DEFRA (2007) How to use the Shadow Price of Carbon in policy appraisal

DTLR (2002) Neighbourhood regeneration: Lessons and evaluationevidence from ten single Regeneration Budget case studies.

English Partnerships (2001) Employment Densities – A Full Guide

HM Treasury (2003) Appraisal and Evaluation in Central Government (known asThe Green Book)

HMSO (1987) Evaluation of the Enterprise Zone

HMSO (1995A) Evaluation of the Enterprise Zone

HMSO (1995B) Evaluation of the Enterprise Zone

HMSO (1995) Evaluation of DTI Funded TEC Services in Support of Smalland Medium Sized Businesses

ODPM (2003) Assessing the Impact of Spatial Interventions: Regeneration,Renewal and Regional Development (known as the 3Rs)

Regeneris (2006) Evaluation of Objective Two Business Support Activities in theNorth East

Rhodes, J and other (1994) English Partnerships: Evaluation Evidence, Parameter Valuesand Expenditure Weights. A Reporrt Prepared for EnglishPartnerships. (September, 1994 – unpublished)

Scottish Enterprise (2007) Economic Impact Assessment Guidance Note

SQW (2007) Interim Evaluation of the Coalfields Regeneration Programmesin England – Final Evaluation Report

Bibliography

Bibliography


Bibliography44

Changes since the second edition(September 2004) of this Guide

There are four principal changes since thesecond edition of this Guide as follows:

(i) Assessing additionality in qualitativeterms

Three approaches are included in thisupdated Guide showing how the qualityof an output/outcome can beincorporated within an additionalityassessment.

(ii) Updated evaluation evidence

A number of more recent evaluationresults have been included within thisupdated Guide. Older sources havebeen replaced where appropriate.

(iii) Tourism interventions

In the previous editions of this Guide,there was no example of a tourismintervention in Section 5. A tourismintervention has been added in thisedition.

(iv) Gross to net additionality ratios

Greater emphasis is given to a simplifiedapproach, which may in somecircumstances be appropriate where anindicative assessment of additionality isrequired.


Appendix A

Appendix A45

Appendix A Major changes since thesecond edition of this Guide


Appendix B

46

Illustrations of additionality in thecontext of different reference cases

This appendix provides illustrations of threedifferent reference cases. The first caseinvolves an improving reference case wherethe intervention involves an additionalimprovement. The second case involves adeteriorating reference case, but where theimprovement associated with theintervention is sufficient to deliver a netoverall improvement. The third case is oneshowing a deteriorating reference case,where the intervention partially offsets thisdeterioration, but not totally – leading to anet deterioration – but not as large as theone that would have occurred without theintervention.

Case 1 – improving reference case

Appendix B

Appendix B Illustrations of additionality in the context ofdifferent reference cases

A = Observed/expected change

B = Increase in Y under thereference case

C = Additional impact ofintervention (e.g., jobs created)

A

B

C

Y(e.g., jobs)

Start – Base Year Time Finish – End State

Baseline


Appendix B

47

Case 2 – deteriorating reference case with overall improvement

Case 3 – deteriorating reference case with overall deterioration

Appendix B


B = Decrease in Y under the reference case

C = Additional impact of intervention (e.g., jobs safeguarded)Y

Baseline



B = Decrease in Y under the reference case

C = Additional impact of interventionOf which:• B = negative impact prevented

(e.g., jobs safeguarded)• C – B = positive impact generated

(e.g., jobs created)

Y


C

C

A

B

B

A

Baseline


Appendix C

48

Interventions optionsThe public sector intervenes to achievespecific objectives and generate particularoutputs and outcomes in a specified periodof time. There will always be a number ofalternative options or ways in which thepublic sector might intervene. As a minimumthese will include: different timings;increasing or decreasing the scale ofinvestment; increasing or decreasing thequality of the outputs, and varying thedelivery arrangements. Whilst not all of thesewill be feasible options in each case, veryfew, if any interventions could not bedelivered in a different way and still meet themajority or all of its objectives.

The identification and assessment ofalternative intervention options is central toappraisal. Without a proper assessment ofthe options it will be difficult to haveconfidence in any assessment of the valuefor money of the intervention. Comparingthe intervention option alone with thereference case will tell you only about theadditionality of that option, it will not tell youwhether or not greater additionality andmore value for money could be achieved bydelivering the intervention in a different way.

It is usual to start the process by generatingand reviewing an initial list (long-list) ofintervention options. In drawing up the initiallist it is good practice to consult those whoare the intended beneficiaries, others whohave experience in delivering similarinterventions and internal or externalexperts. Where this initial list is too long tomake appraisal of all options possible, thelist can be reduced to a shorter-list usingappropriate criteria. The short-listing criteriacould include constraints of a physical, legalor planning nature that make theintervention not feasible or it could be basedon an analysis that showed some optionswere better than others at producing thesame or more outputs at less cost.

The options on the shorter list – whichdepending on intervention size or natureshould include at least four options – willthen be subject to detailed appraisal. Thefull range of outputs and outcomes of theshort-listed option need to be identified,then the level and timing of them estimated.

Appendix C

Appendix C

Crowding out andcrowding in(i) Definition

Crowding out – The tendency for outputs(other than those that increase the rate ofcapacity growth through a supply sideimprovement) to be entirely offset becauseof macro-economic adjustments

(ii) Examples of crowding out

Crowding out is a specific case of an impactthat needs a slightly different type ofconsideration. In assessing additionality thereshould be some recognition of theGovernment’s overall macro-economic policywhich determines the overall level of demandin the economy. That policy is currently aimedat ensuring that, over the longer term,expansion of demand is broadly in line withincreases in capacity. This implies thatboosting prosperity is a matter of raising therate of capacity growth, through supply sideimprovements, as this will allow demand tobe expanded without generating inflation.Thus, increasing economic activity throughprogrammes and policies that do notincrease capacity growth will be totally offsetelsewhere in the economy, either directlythrough displacement or indirectly throughhigher taxes, interest rates and wages as aconsequence of the extra governmentexpenditure. This indirect impact is calledcrowding out and it implies that if there is nosupply side impact associated with anintervention then no additional outputs will begenerated at the national level.

(iii) Factors affecting the scale ofcrowding out

This will largely be determined by the extentto which the impacts arise as a result ofsupply side improvements. Where a specificimpact arises as the result of a supply sideimprovement there is no reason to expect

that there will be a compensating macro-economic adjustment. This will need to betaken into account when an intervention hasa range of impacts only some of which havean impact on the supply side of theeconomy.

(iv) Approaches to estimating

Crowding out is of most relevance in relationto impacts at the national level and hence adegree of proportionality must be applied intaking the effect into account. Typically it willbe more important in the case of policiesand programmes. Large interventions, orinterventions where the costs and benefitsare finely balanced, however may also needto consider the effect. It may be assumedthat:

For areas covering up to 5% of the UKpopulation, crowding out can reasonablybe ignored; and

For areas of 5%-20% of the UK workingpopulation31, it is reasonable to presentresults without taking account ofcrowding out so long as this is explicitlystated.

For programmes covering more than 20% ofthe UK working population, explicit accountshould be taken of crowding out, andestimates of net outputs reducedaccordingly. It should be assumed that inthe absence of a demonstrated supply sideimprovement, crowding out is 100% at thenational level.

Crowding In

It is also possible that an intervention mightresult in crowding in effects, wherebyvariables in the economy adjust and result inan increase in private expenditure. Thus,investment in the physical and humancapital infrastructure might result in a moreefficient level of activity and therefore crowdin, rather than crowd out, private sectorinvestment. The public sector can thuscreate the conditions for increased privatesector activity.


Appendix D

Appendix D49

Appendix D

30 Population is used here as a proxy for the magnitude of the impact. This will be relevant to some but not allinterventions. An alternative would be the share of UK GDP.


Appendix E

50

Housing andadditionality1.0 Introduction

This appendix presents further examples ofhow to assess the additional impact ofhousing programmes and projects. It alsoidentifies some of the key sources ofinformation needed to assess additionality inrelation to housing.

The appendix includes examples of housinginterventions within both low demand andgrowth areas. Like the rest of theAdditionality Guide the focus in thisappendix is on appraisal and therefore anex-ante assessment of expected impacts.

2.0 Additionality in low demandand growth areas

The additionality framework needs to beapplied with due care and consideration indifferent situations – with thought given tothe rationale/objectives for intervening andthe market and policy contexts. Thus, forexample, interventions that lead toincreasing house prices in low demandareas could be seen as a positive marketdevelopment, but would probably benegative in a growth area.

In relation to low demand areas, the focuswill often be on changing the scale andnature of demand within a specific location.As such, supply-side interventions areundertaken in order to stimulate increaseddemand by enhancing the attractiveness ofthe area.

Conversely, in growth or high demand areas,interventions may be concerned withincreasing affordability – for example, byincreasing the supply of affordable or keyworker homes to meet existing unmetdemand, thereby increasing householdnumbers. There may be an increase inhousehold formation in the interventionoption and care must be taken whenassessing displacement. New householdsthat would otherwise not have existed wouldnot be displaced. They may also result inother wider benefits, such as reducing travel

distances, by altering commuting patternsand allowing people to live closer to work.The rationale for intervening will often beabout ensuring sufficient local supply of keyworkers, such as teachers and nurses.

Whilst the additionality framework can inprinciple handle these differences, it willneed to be applied flexibly and thoughtfully –not as a template to be imposedmechanically.

3.0 Examples of how to assessthe additionality of ahousing programme orproject

3.1 Programme level additionality

In developing a programme, additionalitycan be considered through:

(i) an overall assessment of expectedchanges in conditions (the macro ortop-down approach) under differentscenarios; and

(ii) assessing each of the factors(leakage, displacement/substitution,multiplier effects and deadweight) inturn, based upon aggregatingprogramme/project level activities (themicro or bottom up approach). Thisinvolves making evidence-basedjudgements and being explicit aboutthe assumptions for each additionalityfactor and the expected implicationsof a set of actions.

The macro approach would typically bebased upon trend analyses or forecastingmodels, again comparing reference andintervention options. Model-basedapproaches can allow the potentialrepercussions off the direct causal chain tobe considered. However, in many casesproducing such forecasts is notstraightforward and will involve subjectivejudgements about underlying assumptions.A wide range of indicators could potentiallybe assessed, including:

Changes in stock and changes in thestock in long-term or sustained demand,that is housing units for which there isexpected to be long-term demand. A

Appendix E

Appendix E

distinction needs to be drawn betweenoccupation demand (from owners and forrent) and non-occupation demand (buy torent and speculative). The latter will havea price effect, but may or may not resultin occupation of the stock. In most casesit is the former which will be the primaryconcern;

Changes in the condition of the stock;

Changes in the relative price of the stockand total stock value;

Changes in rates of out-migration (as ameasure of resident satisfaction);

Changes in the tenure mix (again as ameasure of social change and willingnessto buy and thus confidence or theaddition of a new type of dwelling, e.g.low cost home ownership to meet adefined need);

Levels of private sector investment in thestock; and

Supporting socio-economic indicators.

In undertaking a micro-based approach theexpected trends in key variables, such aschanges in household formation, will stillneed to be considered. The appropriateoutput/outcome indicators will need to beidentified for any given intervention, but therange of indicators will be the same asthose for the macro-approach. For mosthousing projects, a combination ofindicators may need to be considered,which could be drawn from the following,depending on the objectives of the project:

(i) housing units which are insustained demand. Care needs tobe taken because housing unitscan differ significantly in their scaleand nature – for example, a onebedroom flat and a five-bedroomdetatched house. The appropriatetype of housing unit for any specificproject should be determined byreference to the policy objectivesand the rationale for intervening;

(ii) sustained change in housing stock

value – such effects can be verydifficult to measure on an ex-antebasis and considerable care willneed to be taken in using thisindicator;

(iii) condition of the stock; and

(iv) wider impacts – including localenvironment and quality of life.

In each case, information and evidencewould need to be provided concerningamongst other things:

• Key assumptions – the rationale forthe assumptions used in assessingadditionality. This would normallyinclude information about keymarket segments and specificareas or neighbourhoods;

• Phasing and timing issues –consideration of the likely effectsover time;

• Sensitivity analysis – considerationof the effect of varying keyassumptions. This would be linkedto key market drivers and to therisk assessment for the programmeor project; and

• Monitoring – details of themonitoring framework, including theindicators to be tracked, that will beestablished to monitor additionalityand, in particular, displacementeffects.

3.2 Intervation level additionality

Growth area examples

(i) New build housing project

In developing a project, as with aprogramme, additionality can be assessedby considering each of the factors in turn.For example, for an intervention involving thedevelopment of 120 new housing units on acleared site, of which 30% will be affordable,for which there is believed to be long-termdemand, the following factors would needto be considered at the growth area level:

Reference case (deadweight) – for thisexample, if it is expected that 20% of theunits developed on the site would havebeen affordable anyway in line with thelocal plan requirement. The originalscheme would have been constructed ata lower density, with some 100 units likelyto have been delivered. Thus, 20affordable units would be the gross directeffects under the reference case.

Leakage – if non-target households32 (i.e.those earning more than the minimumlevel specified) were to occupy any of theaffordable units then leakage wouldoccur. However, only those people on theCouncil’s list would be allowed to occupythe affordable homes and thereforeleakage would be zero in both thereference and intervention case options.

Displacement/substitution – if demandfor the units is expected to come fromoutside of the local area or fromhouseholds that would not otherwiseexist then no displacement would occur.In this case, there is excess demand foraffordable housing and, as such, nodisplacement.

Multiplier effects – since the focus ofthis analysis is on housing units then thisfactor is not relevant in this case.

Table E1 shows how the example would beworked through to calculate interventionlevel additionality in terms of the number ofaffordable units in sustained demand.


Appendix E

Appendix E

51

32 Care needs to be taken when assessing leakage in relation to a sub-set of households within target area. Changes in relation to the target households may have positive or negativeeffects on other households within the target area.

The total net additional local effect of theintervention will be 16 affordable residentialunits in sustained demand (the total neteffect for the intervention option minus thetotal net effect for the reference case).

In this case the intervention is not expectedto result in the stimulation of additional,unassisted developments, although somestimulation of confidence is expected.However, it will result in a range of widerbenefits that will also need to be consideredin the appraisal. These might includeenvironmental benefits as a result ofremoving an eyesore, as well asdemonstration effects that will help tochange developer and occupier perceptionsof the area.

(ii) Major brownfield housingdevelopment in a growth area

A major housing scheme, comprising 600homes with 40% key worker and associated

employment and leisure uses, is proposedon a brownfield site within a growth areawith public support. The site is locatedwithin walking distance of the city centre.The scheme will include high design andother standards. As a result of the groundconditions, the affordable housingcomponent and high standards the schemeneeds public sector funding. Without theproject, the less contaminated portion of thesite, which is relatively easily developedwould be expected to come forward,providing 200 homes.

The intervention is designed to helpaccommodate the significant growth inhouseholds that is projected in the growtharea and specific criteria will be imposed inselecting potential occupiers for the keyworker housing for which there is also verysubstantial demand – as such leakage isexpected to be zero. In terms ofdisplacement there is unmet demand and,

as such, product market displacement isalso zero. However, factor marketdisplacement, as a result of the limitedavailability of construction resources isexpected to occur. The level of factormarket displacement is estimated to be10%.

Table E2 shows an assessment of the totalnet additional local housing units insustained demand.

Low demand area examples

(i) Mixed use refurbishment

In this example, it is assumed that publicsector support has been requested towardsa project on a canal-side former millcomplex in a low demand area that willdeliver a mixture of commercial developmentand housing in a variety of new andrefurbished buildings. The Mill complex isGrade II listed.


Appendix E

Appendix E

52

33 By bidding up prices.

Table E1: Example 1 – Affordable housing in growth area

Intervention option Reference case

A Gross direct effects 36 20

B Less leakage from target group/area 0 0


D Less displacement / substitution 0 0


F Plus multiplier effects - -

G = E+F Total gross local effects 36 20

H = G (intervention options) - G (reference case) Total net additional local effect 16

Table E2: Example 2 – Key worker houses in growth areas (housing units in sustained demand)



B Less leakage from target group/area 0 0


D Less displacement (factor33and product 60 20market) / substitution


F Plus multiplier effects - -


H = G (intervention options) - G (reference case) Total net additional local effect 360

The target area for the intervention is thesite itself and the wider regeneration area. Itis expected that the project will deliver 50housing units and 2,000 sq m ofcommercial floorspace (assumed to be B1office space). Without public sector supportit is likely that one of the former millbuildings on the site, which is in the beststate of repair, would be brought forward bythe private sector, and would deliverapproximately 20 housing units and 500 sqm of commercial floorspace. It is notexpected that the project will cause a largedecrease in the number of housing unitsbuilt elsewhere in the target area sincedemand is low, although because of thenature and quality of this scheme sustained

demand is anticipated. As such adisplacement rate of 20% has beenassumed.

The objective of the public sector supportingthe housing component of this project is togenerally attract new residents to the area.Thus ‘leakage’ of housing benefits is notconsidered an issue in this case. However,jobs taken be people outside of theregeneration area has been assessed.

Table E3 sets out an assessment of the netadditional housing units in sustaineddemand generated by the proposedintervention option at the level of the siteitself and at the target area.

Table E4 considers the net additional

employment outputs at the target area level.The site level has not been considered inthis case since the focus is on creating localemployment opportunities.

Table E5 considers the overall effects ofhousing and employment at the site andtarget area levels.

(ii) Student housing

Another example of how project leveladditionality can be assessed using a morequalitative approach is set out in Table E6. Inthis case, the key output and outcomeareas to be considered are housing andeconomy. The analysis includes discussionof both supply and demand sidedisplacement.


Appendix E

Appendix E

53

34 There may be multiplier effects of the expenditure associated with the housing construction, the’ not applicable’ refers to the likelihood that this spend will not result in more housingunits being built.

Table E3: Example 3 – Housing in low demand area (housing units in sustained development) – at site and target area levels

Intervention Option Reference Case AdditionalitySite level Target Site level Target Site level Target area level area level area level

A Gross direct housing units 50 50 20 20

B = Not applicable Estimated leakage N/A N/A N/A N/A

C = A-B Gross local direct effects 50 50 20 20

D = estimated displacement Displacement – 20% 0 10 0 4

E = C-D Net local direct effects 50 40 20 16

F = Not applicable Multiplier N/A N/A N/A N/A

G = E+F Total gross local effects 50 40 20 16

H = G (intervention option) Total net additional local effects 30 24- G (reference case)

Table E4: Example 3 – Employment in low demand areas – target area levels


A (@ 20 sq m per job) Gross direct employment 100 25

B = estimated leakage Estimated leakage – 20% (Jobs taken up 20 5by people resident outside of the target area)

C = A-B Gross local direct employment 80 20

D = estimated displacement Displacement – 40% 32 8

E = C-D Net local direct employment 48 12

F = Combined income & supply multiplier Multiplier at 1.2 10 2


H = G (intervention option) Total net additional local effects 44- G (reference case)


Appendix E

Appendix E

54

Table E5: Example 3 – Summary of housing and employment additionality – at site and target area levels

Intervention Option Reference Case AdditionalitySite level Target Site level Target Site level Target area level area level area level

Gross direct housing units in sustained demand 50 50 20 20 50 50

Gross to net housing additionality effect 100% 80% 100% 80% 60% 48%

Total net additional local housing effects 50 40 20 16 30 24

Gross direct employment 100 100 25 25 100 100

Gross to net employment additionality effect - 58% - 56% - 44%

Total net additional local employment effects - 58 - 14 - 44

Table E6: Example 4 – Student housing in low demand area

Programme: Student housing

Intervention option: Subsidised refurbishment of 100 vacant housing units35 for students by the local University

Reference case: Ad hoc refurbishment of 20 units based on past trends

Leakage

Displacement

Economic multipliers

Deadweight

Net additionality

Housing

• No leakage due to project design (i.e. units exclusively for students).

• It is estimated that 10 of the students would have occupied private rentedaccommodation in the low demand area. However, much of this is in a poorstate of repair and would give students a lower standard of living.

• Issue about potential future vacancies in private rented stock. Linked to privaterented programme which forms another strand of low demand arearegeneration activity.

• Students may be disruptive and cause other residents to move away. Carefulchoice of units and design of works is required.

• Students would have found accommodation in wider area (say sub-region),therefore the vast majority of demand will be displaced at this level (i.e. it isexisting demand rather than new at the sub-regional level).

• Not relevant because the focus is on housing outputs and outcomes.

• 20 units would be expected to be provided anyway through refurbishment. Theremaining 80 units would, on the basis of past trends, have remained vacant.

• Likely to result in additional quantitative and qualitative benefits at the local, lowdemand area level. Careful choice of units and design will be needed to ensurethat student accommodation does not result in disruption to existinghouseholds.

• The majority of the students would not have been living in the low demand area in the absence of this programme.

• At the wider sub-regional level the demand would have arisen anyway andtherefore the activity in a quantitative sense would not be net additional.However, there would be qualitative benefits in terms of the standard ofaccommodation. The areas from which demand is likely to be displaced aresignificantly more buoyant than the target low demand area. Wider level, netadditional benefits would include positive environment affects and increasedvitality of the area. Vacancy rates would be reduced.

Economy

• Students spend outside area.• Issue of supply of local facilities to meet demands from

students (e.g. local fast food restaurants). Furtherinvestigation and possible linked programme.

• Reduction in private rented/room letting and theassociated with this.

• Students cause other residents to move away resulting in a loss of local expenditure.

• At the wider level, the vast majority of this economicactivity would have happened anyway since the studentswould have been spending their student loans elsewhere.

• Purchases by local shopkeepers due to increased turnover and increased local spending as a result ofgreater local incomes.

• Economic activity associated with the 20 units.

• Again net additional local benefits would be expected,although the availability of appropriate facilities will requirefurther consideration.

• At the wider level the majority of the student relatedeconomic activity is likely to be displaced.

35 Care needs to be taken to ensure that where housing units arebeing considered that the nature of the units delivered undereach scenario is comparable. Where this is not the case anddifferent market segments are being targeted, then significantcare needs to be taken in assessing additionality and inundertaking the project appraisal more generally.

Examples of how to use the readyreckoners in housing projects

An intervention is being assessed at a growtharea level and involves the provision of 50starter homes. In the absence of theintervention it is estimated that only five starterhomes would otherwise come forward in thearea. However, some of these newhouseholds would have been expected to findlocal accommodation in the form of older,poor quality stock. Thus, the interventionwould result in qualitative benefits. Based on

local market information and answering thequestions set out in the main AdditionalityGuide, the expected additionality effects inrelation to the number of units in sustaineddemand at the growth area level areestimated as set out in Table E7.

Table E8 presents a quantitative assessmentof additionality for the starter homesintervention at the growth area level.

Another example would be an interventionto promote executive homes within a low

demand area. In this case, the level ofadditionality, at the low demand area level,would be significantly higher since very littleof this activity would have taken placeanyway. In this example, it is assumed that100 executive homes would be constructedunder the intervention option and noneunder the reference case. For thisintervention, the expected additionalityeffects, based on local market analysis andland use planning allocations at the lowdemand area level, are as in Table E9.


Appendix E

Appendix E

55

36 By bidding up prices.

Table E7: Ready reckoner assumptions – starter homes in growth areas


Leakage Low – 10% Low – 10%

Displacement/substitution Medium – 50% High – 75%

Multiplier effects N/A N/A

Table E9: Executive Homes – Low Demand Area level


Leakage None – 0% None – 0%

Displacement/substitution None – 0% None – 0%


Table E8: Ready reckoner worked example – starter homes (growth area level) – units in sustained demand



B Less leakage from target group/area – i.e. homes going to 5 1none target group (Intervention – 10% and reference case – 10%)


D Less displacement (factor36 and product market) substitution 23 3(Intervention – 50% and reference case – 75%)


F Plus multiplier effects N/A N/A


H = G (intervention option) Total net additional local effects 21- G (reference case )

As a result, the total number of netadditional executive homes would be 100 at the low demand area level (i.e. 100 less zero).

However, the level of additionality of theexecutive homes intervention would besignificantly lower at the sub-regional level.The project would be expected to draw

demand from both the city centre and sub-urban areas, where demand is high. Theestimated level of additionality, based uponmarket analysis, are as follows:

Table E11 presents the results of theassessment of additionality at the sub-regional level.


Appendix E

Appendix E

56

Table E10: Executive homes – sub-regional level


Leakage None – 0% No homes built in area

Displacement/substitution High – 75% No homes built in area


Table E11: Ready reckoner worked example – Executive homes (low demand, sub-regional level) – units in sustained demand



B Less leakage from target group/area 0 0(Intervention and reference case – 0%)


D Less displacement (factor and product market) 75 0substitution (Intervention – 75% and reference case – zero )


F Plus multiplier effects N/A N/A

G = E+F Total net local direct effects 25 0

H = G (intervention option) Total net additional local effects 25- G (reference case)

4.0 Information sources

A wide-range of data will need to beconsidered to assess additionality in relationto a housing intervention.

In order to make informed decisions aboutwhat intervention will minimise negativeeffects and to ensure that the maximumadditional benefits are being delivered,information will be required on a range ofsubjects, including potentially:

(i) Population, migration, householdand economic forecasts – theseforecasts should include informationabout areas within and outside theintervention area boundary. They providethe framework within which theassessment of additionality will beundertaken. The economic, social andother factors that have been assumed to‘drive’ or underpin the forecasts shouldbe explained and the likely housingimplications considered. Whereverpossible forecasts should relate to thespecific areas of impact underconsideration and consider the type ofhousing demanded. Alternative futurescenarios will often need to beconsidered to test the robustness of theproposed programme or project.

(ii) Policy context and other publicsector initiatives – in particular,planning, housing and economicdevelopment policies will need to bereviewed. These will include the RegionalHousing Strategy (RHS), RegionalSpatial Strategy (RSS) and RegionalEconomic Strategies (RES), as well aslocal policies, such as LocalDevelopment Framework (LDF) and, inparticular, housing policies. In addition,other public sector initiatives, inparticular those concerned withregeneration, education and health willneed to be considered.

(iii) Existing and forecast housing landsupply – this should be based uponexisting and proposed RSS, LDF andother housing policies, together with anassessment of outstanding planningpermissions, within the various areas of

impact and broken down into marketsegments (including type of dwelling andtenure). Alternative housing land supplyscenarios may again need to bedeveloped and tested.

(iv) Housing market conditions – this willinclude an analysis of current conditionsand past trends for the local and widerarea. The types of data that will need tobe considered include: house prices;land prices; rental levels; stock by typeof dwelling and tenure; sale periods;turnover within the stock; vacancy ratesby type and void information andhousing management data for socialrented accommodation. It will often benecessary to assemble and analyseneighbourhood level data to understandhow the housing market is operatingwithin a local area.

(v) Affordability – Housing Needs Studies,income and house price data,information from key local public sectoremployers, e.g. Police and NationalHealth Service (NHS).

(vi) Surveys – these are likely to be a keysource of data in assessing leakage,displacement and deadweight.Information may be collected from anumber of primary sources:

Individual/households – can bequestioned about their views, likely futureneeds, aspirations, attitudes and locationdecisions.

Direct questioning of developers – ontheir expected behaviour and on theirproposed approach to marketing andexpected sources of demand.

Mover/beneficiary survey – occupiers ofrecent developments, supported projectsor in/out movers, can be questionedabout where they moved from, why andwhat they would have done if the homesthey moved to were not available.

Focus groups – detailed discussions withgroups of, for example, residents orrecent movers can be useful in providingin-depth discussions about complexresidential choice decisions.

Other specific surveys – for example,housing chain surveys to determine indetail the nature of housing movementswithin and beyond the local area.

Business views, again obtained usingsurveys.

(vii)Monitoring data and evaluationresults.

In each case the analysis should considerthe likely timing of changes and theseshould be related to the nature and phasingof the intervention.

In relation to displacement there are anumber of specific indicators that may needto be assessed and then subsequentlymonitored, depending on the objectives ofthe intervention, these may include:

(i) Housing stock, mix and condition/standard

Tracking changes in the total number andnature of the stock will be important in orderto assess potential displacement effects.The information assembled should includedata about the mix of housing (number ofbedrooms, flat or houses), as well as thecondition of the stock.

(ii) New starts and completions

The level of new development activityunderway and completions should also bemonitored. Identifying trends in activitywithin and outside the area can help toinform judgements about displacement.

(iii) Occupancy – Hard to let/vacancy/voidrates

Vacancy is a further key indicator, andmonitoring of adjacent areas will indicatewhether problems being tackled in an areaare being displaced to adjacent areas.Registered Social Landlords (RSL) are ableto provide indicators of hard to let premises.Void rates is another useful indicator.

(iv) House prices and rental levels

The issue here is the extent to which theyare affected by the programme or project,taking into account general trends. PropertyAgents, financial institutions, the DistrictValuer and RSL’s are good sources of thisinformation.


Appendix E

Appendix E

57

(v) Tenure patterns

Another key indicator is tenure mix, typicallyowner occupied, private rented, and socialhousing. Changes need to be monitoredagainst typical conditions in the area itselfand in other areas. Displacement oftentakes the form of ‘gentrification’ whererelatively more affluent buyers force outlower income residents, who may be forcedto move elsewhere.

(vi) Houses of Multiple Occupation (HMOs)

HMO’s are an important indicator ofpressures at the lower end of the privaterented sector. Often, these will be displacedfrom the intervention area, and move toadjacent areas, potentially causing socialand environmental nuisance.

(vii) Turnover

Turnover can be an indicator of vitality in ahousing market, but high levels could be anindicator of displacement. However, it mightreflect the transitory nature of the areas andtheir populations. As such careful analysisand cross-checking with, for example,survey results will often be needed ifindicators like turnover are to be correctlyinterpreted.

(viii) Overcrowding

A further possible indicator, which may berelevant in a limited number ofcircumstances, is the degree ofovercrowding where housing demandincreases and housing supply is unable tocope.

(ix) Density/number of units

Density (and therefore the total number ofunits in an areas – in particular, occupiedunits) is again a possible factor indisplacement, Typical low demand areaterraced properties are a very efficient use ofland, achieving 50-60 dwellings per hectare.Replacement dwellings may be at lowerdensity (say 30 dwellings/hectare plus),which could also affect local businesses,especially shops and local services.

(x) Building costs

Engineering the supply of housing in a localarea will attract development activity,involving both main contractors and sub-contractors of various types. There may wellbe a tendency to ‘bid up’ prices, as well asto displace activity from other, non-supported surrounding areas. A measure ofthis would be to monitor local building costinflation, which local quantity surveying firmscould assist with.

(xi) Community

Many communities have remained togetherin lower cost housing because of the abilityto house extended families in closeproximity, and because they may not wish totake traditional forms of mortgage.Acquisition and re-housing of such familiescan bring significant displacement ofcommunities to other adjacent areas, orindeed between urban areas in differentparts of the sub-region. These displacementeffects could be identified through surveys.


Appendix E

Appendix E

58

Contributors

Steve Carr, English Partnerships

Claire O’Shaughnessy, English Partnerships

Simon Dancer, English Partnerships

Graham Russell, AMION

in association withEdward Palmer, CLG.

Al= [Gl x (1-L) x (1-Dp) x (1-S

(1-L*) x (1-Dp*) x (1-S*) x M*]

*) x (1-Dp*) x (1-S*) x M*]

Dp) x (1-S) x M] – [Gl*x (1-L*) x (1-Dp*) x (1-S*) x M*]

Al= [Gl x (1-

A l= [Gl x (1-L) x (1-

www.englishpartnerships.co.ukPublication date: October 2008 Ref: ENG0046

English Partnerships is able to provide literature inalternative formats including large print, braille andaudio. Please contact us on 01925 644741 or byemail at [email protected] for furtherinformation.

For further information please contact:

Simon Dancer

Economist

English Partnerships

110 Buckingham Palace Road

London

SW1W 9SA

T 020 7881 1634

F 020 7881 1666

E [email protected]

Documents

Additionality Guide · (referred to as ‘The Green Book ... in any detail in this Guide.7 English Partnerships | Additionality Guide Third Edition| October 2008 Introduction 1 Introduction