Docstoc

080421 IA merged and final-2

Document Sample
080421 IA merged and final-2 Powered By Docstoc
					                                         Summary: Intervention & Options
 Department /Agency:                                   Title:
 Department for Transport                              Impact assessment of proposed changes to car driving
                                                       training and testing


 Stage: Consultation                                   Version: 1                                           Date: 7 May 2008
 Related Publications: Learning to Drive Consultation paper; Learning to Drive: the evidence


 Available to view or download at:
 http://www. www.dsa.gov.uk/learningtodrive
 Contact for enquiries on impact assessment: Kaite Emmerson                                   Email: kaite.emmerson@dft.gsi.gov.uk

 What is the problem under consideration? Why is government intervention necessary?
 Introduction of more effective training and testing system for car drivers aims to prepare drivers better for
 independent driving.There is evidence that learner drivers have an unstructured and poorly integrated
 approach to learning to drive. New qualified drivers emerge with training and testing with significant gaps in
 experience and a poor understanding of road safety. They exhibit a range of unsafe driving attitudes and
 behaviours. One in five new drivers will have an accident of some sort within a year of passing their states.
 Young and new drivers are vulnerable to being involved in casualty accidents and therefore pose a danger
 to themselves, their passengers and other road users. This presents a considerable challenge in terms of
 the Government's road safety strategy.
 What are the policy objectives and the intended effects?
 The aim of the proposed intervention is to reform the current driver training and testing regime to ensure that
 people seeking to obtain a full driving licence are fully equipped to drive independently and safely at all
 times. The intended effects are to equip and educate new drivers with the skills, knowledge, attitudes and
 behaviour to be safe life-long drivers; to improve their driving and by extension to reduce the casualty rate
 for newly qualified drivers. The policy objectives are also to increase the efficiency and effectiveness of the
 training and testing process.



  What policy options have been considered? Please justify any preferred option.
 At this stage the impact assessment considers five options: option A - do nothing; option B - changes to
 driver training and testing system; option C - restrictions on learner drivers; option D - restrictions on newly
 qualified drivers; and option E - combination of options B and D.
 The preferred option is B, which represents a combination of voluntary education and mandatory testing
 arrangements to encourage drivers to encompass a broader range of knowledge, skills, attitudes and
 behaviour to driving on the roads.


When will the policy be reviewed to establish the actual costs and benefits and the achievement of the desired
effects? It is anticipated that implementation of new initiatives will take 2-3 years. Initial evaluation of new tests
will be done after 3 months, a more full assessment will made after 12 months and an evaluation of the
programme after five years.
 Ministerial Sign-off For consultation stage Impact Assessments:
       I have read the Impact Assessment and I am satisfied that, given the available evidence, it
       represents a reasonable view of the likely costs, benefits and impact of the leading options.
 Signed by the responsible Minister:
 Treat as signed
 ............................................................................................................ Date: 30 April 2008


                                                                                                                                     1
                                       Summary: Analysis & Evidence
Policy Option: B                      Description: Changes to Driver Training and Testing



                  ANNUAL COSTS                  Description and scale of key monetised costs by ‘main
                                                affected groups’ The potential major costs relate to ongoing costs of
           One-off (Transition)         Yrs     delivering pre-driver education training and the distribution of
           £ 30m-54m                     2      workbooks for learner drivers. Potentially, the introduction of new test
                                                arrangements may incur additional DSA operating costs (if so this will
COSTS




           Average Annual Cost                  fall to candidates).
           (excluding one-off)

           £ 56m                                                                 Total Cost (PV)        £ 509m-533m
           Other key non-monetised costs by ‘main affected groups’ Voluntary pre-driver education would
           involve opportunity costs for individuals, if only, in terms of forgone leisure activities.



                ANNUAL BENEFITS                 Description and scale of key monetised benefits by ‘main
                                                affected groups’ Improved driver training and testing is expected to
           One-off                      Yrs     lead to improved driving and an associated reduction in casualty and
           £0                                   damage only accidents amongst newly qualified drivers and their
                                                passengers. More structured learning is expected to increase pass
BENEFITS




           Average Annual Benefit               rate resulting in cost savings for learner drivers in the form of fewer
           (excluding one-off)                  lessons and tests.
           £ 293m-749m                                                         Total Benefit (PV)       £ 2523m-6451m
           Other key non-monetised benefits by ‘main affected groups’ Better driving habits are expected to be
           deliver more fuel efficient driving and fewer CO2 emmissions. The reduction in accidents would be
           expected to deliver less congestion for other road users.


Key Assumptions/Sensitivities/Risks Impacts depend to a large extent on scenarios, which are based on
limited information or on evidence available from elsewhere. This may not be transferable to the British
context. Therefore, estimates may need to be revised when further information/clarification is obtained
through consultation.

Price Base               Time Period         Net Benefit Range (NPV)                      NET BENEFIT (NPV Best estimate)
Year 2006                Years 10            £ 1990m to 5942m                             £ 3966m

What is the geographic coverage of the policy/option?                                                   Great Britain
On what date will the policy be implemented?                                                            In 2-3 years
Which organisation(s) will enforce the policy?                                                          DSA
What is the total annual cost of enforcement for these organisations?                                   £ 0 (as currently)
Does enforcement comply with Hampton principles?                                                        No
Will implementation go beyond minimum EU requirements?                                                  No
What is the value of the proposed offsetting measure per year?                                          £0
What is the value of changes in greenhouse gas emissions?                                               £ not quantified
Will the proposal have a significant impact on competition?                                             No
Annual cost (£-£) per organisation                                     Micro            Small           Medium             Large
(excluding one-off)
Are any of these organisations exempt?                                      No               No               N/A              N/A
Impact on Admin Burdens Baseline (2005 Prices)                                                           (Increase - Decrease)

Increase of             £         0          Decrease of     £         0            Net Impact           £           0
                                                 Key:      Annual costs and benefits: Constant Prices        (Net) Present Value

                                                                                                                                     2
                                        Summary: Analysis & Evidence
Policy Option: C                       Description: Restrictions on Learner Drivers



                  ANNUAL COSTS                  Description and scale of key monetised costs by ‘main
                                                affected groups’ Estimates generated to the welfare cost of restrictions
           One-off (Transition)          Yrs    imposed to learner drivers before obtaining full driving licence in
           £0                                   relation to a minimum learning period and raising the full driving age
COSTS




           Average Annual Cost
           (excluding one-off)

           £ 233m-436m                                                          Total Cost (PV)        £ 2004m-3750m
           Other key non-monetised costs by ‘main affected groups’ Costs of enforcement in ensuring
           compliance with proposed minimum learning period, driving age and minimum number of lessons,
           which aredifficult to quantify. Delay in obtaining licence may inhibit employment opportunities for some
           individuals. An estimate may be generated if further information is gathered in consultation.

                ANNUAL BENEFITS                 Description and scale of key monetised benefits by ‘main
                                                affected groups’ Restrictions on learner drivers is expected to realise a
           One-off                       Yrs    reduction in casualty and damage only accidents involving newly
           £0                                   qualified drivers and their passengers.
BENEFITS




           Average Annual Benefit
           (excluding one-off)

           £ 19m-93m                                                          Total Benefit (PV)       £ 160m-800m
           Other key non-monetised benefits by ‘main affected groups’ Benefits of reduced accidents involving
           casualties will accrue to all road users in the form of lower insurance premia and reduced congestion.



                  Key Assumptions/Sensitivities/Risks Welfare cost estimates are subject to debate on the value
                  attached by learners to a driving licence. Limited evidence is available in relation to the effect of
                  this option in reducing the number of accidents by newly qualified drivers. Estimates will need to be
                  revised when further information/clarification is obtained.
Price Base               Time Period        Net Benefit Range (NPV)                      NET BENEFIT (NPV Best estimate)
Year 2006                Years 10           £ -3590m to -1204m                           £ -2397m

What is the geographic coverage of the policy/option?                                                  Great Britain
On what date will the policy be implemented?                                                           To be confirmed
Which organisation(s) will enforce the policy?                                                         Police/CJS/DSA
What is the total annual cost of enforcement for these organisations?                                  £ see evidence base
Does enforcement comply with Hampton principles?                                                       No
Will implementation go beyond minimum EU requirements?                                                 No
What is the value of the proposed offsetting measure per year?                                         £0
What is the value of changes in greenhouse gas emissions?                                              £ negligible
Will the proposal have a significant impact on competition?                                            No
Annual cost (£-£) per organisation                                    Micro            Small           Medium          Large
(excluding one-off)
Are any of these organisations exempt?                                     No               No              N/A            N/A
Impact on Admin Burdens Baseline (2005 Prices)                                                          (Increase - Decrease)

Increase of             £ negligible        Decrease of     £ negligible           Net Impact           £ negligible
                                                 Key:     Annual costs and benefits: Constant Prices     (Net) Present Value

                                                                                                                                 3
                                       Summary: Analysis & Evidence
Policy Option: D                      Description: Restrictions on Newly Qualified Drivers



                  ANNUAL COSTS                 Description and scale of key monetised costs by ‘main
                                               affected groups’ Estimates generated to the welfare cost of restrictions
           One-off (Transition)         Yrs    imposed on newly qualified drivers in the hours they are permitted to
           £0                                  drive and types of passengers they are allowed to carry.
COSTS




           Average Annual Cost
           (excluding one-off)

           £ 300m-563m                                                         Total Cost (PV)        £ 2582m-4842m
           Other key non-monetised costs by ‘main affected groups’ Costs of enforcement in ensuring
           compliance with proposed restrictions on newly qualified drivers, which are difficult to quantify.
           Restrictions on passengers and night time driving may inhibit employment opportunities for some
           individuals.

                ANNUAL BENEFITS                Description and scale of key monetised benefits by ‘main
                                               affected groups’ Restrictions on learner drivers are expected to realise
           One-off                      Yrs    a reduction in casualty and damage only accidents involving newly
           £0                                  qualified drivers and their passengers.
BENEFITS




           Average Annual Benefit
           (excluding one-off)

           £ 109m-219m                                                       Total Benefit (PV)       £ 941m-1882m
           Other key non-monetised benefits by ‘main affected groups’ Benefits of reduced casualties will
           accrue to drivers in the form of lower insurance premia and reduced congestion.



Key Assumptions/Sensitivities/Risks Welfare cost estimates are subject to debate on the value attached by
learners to a driving licence. Limited evidence is available in relation to the effect of the option in reducing
the number of casulaties of newly qualified drivers and their passengers. Estimates will need to be revised
when further information/clarification is obtained.

Price Base               Time Period       Net Benefit Range (NPV)                      NET BENEFIT (NPV Best estimate)
Year 2006                Years             £ -3901m to -700m                            £ -2301m

What is the geographic coverage of the policy/option?                                                 Great Britain
On what date will the policy be implemented?                                                          To be confirmed
Which organisation(s) will enforce the policy?                                                        DSA/Police/CJS
What is the total annual cost of enforcement for these organisations?                                 £ see evidence base
Does enforcement comply with Hampton principles?                                                      No
Will implementation go beyond minimum EU requirements?                                                No
What is the value of the proposed offsetting measure per year?                                        £0
What is the value of changes in greenhouse gas emissions?                                             £ negligible
Will the proposal have a significant impact on competition?                                           Yes/No
Annual cost (£-£) per organisation                                   Micro            Small           Medium             Large
(excluding one-off)
Are any of these organisations exempt?                                    No               No               N/A              N/A
Impact on Admin Burdens Baseline (2005 Prices)                                                         (Increase - Decrease)

Increase of             £         0        Decrease of     £         0            Net Impact           £           0
                                                Key:     Annual costs and benefits: Constant Prices        (Net) Present Value

                                                                                                                                   4
                                       Summary: Analysis & Evidence
Policy Option: E                      Description: Combination of Options B and D



                  ANNUAL COSTS                  Description and scale of key monetised costs by ‘main
                                                affected groups’ As identified in options B and D
           One-off (Transition)         Yrs
           £ 30m-54m                     2
COSTS




           Average Annual Cost
           (excluding one-off)

           £ 356m-618m                                                           Total Cost (PV)        £ 3092m-5375m
           Other key non-monetised costs by ‘main affected groups’ As identified in options B and D




                ANNUAL BENEFITS                 Description and scale of key monetised benefits by ‘main
                                                affected groups’ As identified in options B and D
           One-off                      Yrs
           £
BENEFITS




           Average Annual Benefit
           (excluding one-off)

           £ 298m-653m                                                         Total Benefit (PV)       £ 2562m-4949m
           Other key non-monetised benefits by ‘main affected groups’ As identified in options B and D




Key Assumptions/Sensitivities/Risks As identified in options B and D




Price Base               Time Period         Net Benefit Range (NPV)                      NET BENEFIT (NPV Best estimate)
Year 2006                Years 10            £ -2813m to 1857m                            £ -478m

What is the geographic coverage of the policy/option?                                                   Great Britain
On what date will the policy be implemented?                                                            To be confirmed
Which organisation(s) will enforce the policy?                                                          Police/CJS/DSA
What is the total annual cost of enforcement for these organisations?                                   £ see evidence base
Does enforcement comply with Hampton principles?                                                        No
Will implementation go beyond minimum EU requirements?                                                  No
What is the value of the proposed offsetting measure per year?                                          £0
What is the value of changes in greenhouse gas emissions?                                               £ not quantified
Will the proposal have a significant impact on competition?                                             Yes/No
Annual cost (£-£) per organisation                                     Micro            Small           Medium             Large
(excluding one-off)
Are any of these organisations exempt?                                      No               No               N/A              N/A
Impact on Admin Burdens Baseline (2005 Prices)                                                           (Increase - Decrease)

Increase of             £         0          Decrease of     £         0            Net Impact           £           0
                                                 Key:      Annual costs and benefits: Constant Prices        (Net) Present Value

                                                                                                                                     5
Evidence base (for summary sheets)




                                     1
1 Introduction

This document contains a preliminary impact assessment of the proposed changes to the Great Britain
driver training and testing regime, detailed in the Driving Standards Agency’s Safer Driving for Life
Consultation. More information about this consultation can be found at www.dsa.gov.uk/learningtodrive. The
assessment has been produced by ECOTEC Research and Consulting on behalf of the Driving Standards
Agency (DSA) and the Department for Transport's Road User Safety Division.


1.1   Preparing the impact assessment

The initial assessment of impacts has been prepared based on current understanding and will therefore need
to be revised as areas of uncertainty are clarified and further information is obtained through consultations. It
should be noted that some elements of the options have not yet been costed whilst the impacts
depend to a large extent on scenarios often based on evidence available from elsewhere, which may not
be transferable to the British context.

This document relates solely to the training and testing of car drivers. The values used in the assessment are
based on 2006 prices.

Perhaps the most sensitive assumption applied throughout this assessment relates to the number of drivers
who are actively learning and the estimated number of casualties and damage only accidents involving
newly qualified drivers and their passengers each year (based on 2006 road casualty data). In terms of
casualties and damage only accidents, the numbers vary considerably depending on the assumptions made
and owing to limitations associated with STATS19 data source and the difficulty in eliminating double-counting
from the data, the impact assessment uses the estimated new driver and their passengers casualties as a
baseline rather than all casualties (see Annex A). Further consideration will be undertaken to improve the
methodology relating to estimating all new drivers casualties in the future.

In light of the adopted baseline, the impact assessment recognises that benefits from different options might
represent an underestimation of the reduction in the number of casualties and damage only accidents,
however, a number of other factors may have the effect of overstating the benefits, for example issues relating
to compliance, displacement, and the attribution of blame for accident, i.e. the involvement of a newly qualified
driver in an accident does not necessarily mean that they are cause of it. Although the actual size of impact
depends on the base and compliance and other assumptions, the relative size (i.e. ranking of the
options) is consistent and robust regardless of the base.


1.2   Rationale for intervention

There is a strong case in favour of changes to the current car driver training and testing system. The
introduction of a more effective training and testing system for car drivers aims to prepare drivers better, once
qualified, for independent driving, which in turn would result in a reduction of road casualties and accidents. A
wide range of research indicates that there are issues associated with learning to drive and preparing for
driving tests, gaps in the driver training and testing system and problems with the ability, experience and
behaviour of newly qualified drivers. Statistics show that newly qualified drivers are vulnerable to being
involved in accidents and therefore pose a danger to themselves, their passengers and other road users.
Research estimates that around 1 in 5 newly qualified drivers will have an accident of some sort within one




                                                                                                                    2
year of passing their test. 1 Newly qualified drivers have a higher accident risk, as do their passengers, and
they present a considerable challenge in terms of the Government’s road safety strategy.

The higher accident risk for newly qualified drivers is also reflected in the higher insurance premia which they
face. According to research from the Association of British Insurers (ABI, 2006), young drivers – defined as
those aged 17-20 – who are likely to be newly qualified are twice as likely to make an insurance claim as older
drivers, and the average value of each claim is three times greater. More evidence on why there is a need to
transform driver training and testing can be found in “Learning to drive: the evidence”, available at
www.dsa.gov.uk/learningtodrive.

The Department for Transport (DfT) and the DSA believe that the current driver training and testing regime is
not adequately preparing learner drivers for the challenges of driving unaccompanied post-test, and is
therefore not providing the right level of public service. Good progress has been made by the Department and
stakeholders in moving towards its casualty reduction targets but there is still some way to go. Improving the
driving behaviour and attitudes of learner and newly qualified drivers has been identified as key to securing
further road safety improvements. Currently no part of the training or testing regimes in Britain addresses
attitudes to driving and road safety, although other European countries have started to explore methods of
doing this. There is clear evidence that attitudes to driving are already formed by the time young people begin
learning to drive. A report for the DfT 2 suggests that attitudes are formed during pre-driver stage and have a
corresponding influence when pre-drivers become learners and newly qualified drivers.

Intervention is required as the costs arising from bad driving from newly qualified drivers affect all road users.
It is the Government's responsibility to administer testing and licences, and also to regulate driving instruction
and promote road safety which maximises public benefit in a cost effective manner.


1.3     Objectives and intended impacts

The aim of the proposed intervention is to reform the current driver training and testing regime to ensure that
people seeking to obtain a full driving licence are fully equipped to drive independently and safely at all times.
More specifically, the objectives (and intended effects) are:

          • to equip and educate new drivers with the skills to be safe life-long drivers;
          • to reduce the casualty rate for newly qualified drivers;
          • to raise the standards of trainers and examiners, ensuring that they have the skills to teach the
            required competences 3 effectively and efficiently;
          • to improve education and training arrangements for pre-, learner and newly qualified drivers to
            encompass a broader range of knowledge, skills, attitudes and behaviour (using modern
            competency based systems);
          • to increase the efficiency and effectiveness of the training process;
          • to increase the efficiency and effectiveness of the testing process;
          • to develop modern, effective and fair assessments of driving ability for examiners to use; and,
          • to assist improved compliance by education and assessment arrangements that promote inclusion
            and engagement.

1
  Available at http://www.dsa.gov.uk/Documents/FOI/Statistics/IA00339PA.pdf [accessed 25th January 2008]
2
  Deighton C, & Luther R, (2007) Pre-driver Education: A Critical Review of the Literature on Attitude Change and Development, Good Practice in Pre-
driver Education and Programme Effectiveness Department for Transport: London
3
  Learner drivers will be encouraged to develop competencies that lead to better and safer driving. The driver competencies will be published at the time
of public consultation



                                                                                                                                                        3
1.4   Options to be considered

The options considered here are as follows:

a. Do nothing – i.e. maintain the status quo of the existing training and testing system.

b. Changes to driver testing and training – which represents a combination of options in relation to
   voluntary education and mandatory testing, including:
   • voluntary
      - enhancing pre-driver educational opportunities, available in schools, colleges and other educational
         institutions, and providing recognition of achievement by introducing a foundation certificate of safe
         road use;
      - addressing attitudes early by a combination of assessment and suitable educational intervention,
         such as peer group work;
      - improving the accessibility of learning and assessment.
   • mandatory
      - producing a modern training syllabus to guide professional driving instructors and encourage others
         who supervise accompanying drivers to develop competencies in a more structured approach;
      - empowering learner drivers, giving each a workbook which supports them in working through the
         driving task and enables the student to monitor progress;
      - remodelling the components of the driving test;
      - a modular approach to training and testing involving for example the decoupling of the multiple
         choice theory test and the hazard perception test;
      - a new practical test incorporating independent driving and stronger feedback.

c. Restrictions on learner drivers – to include:
     - a minimum learning period of 12 months;
     - increasing the full driving age to 18 years;
     - reducing the provisional licensing age to 16.5 years; and,
     - a minimum number of lessons with ADIs.

d. Restrictions on newly qualified drivers – to include:
     - limitations on the number of passengers carried (and their age); and
     - a curfew on night-time driving.

e. Changes to driver training and testing and restrictions on newly qualified drivers – this would involve
   a combination of the changes/measures proposed in options b and d.



1.5   Structure of this report

The remainder of this report is structured as follows:

            • Section 2 sets out details of the sectors/groups likely to be affected;
            • Section 3 provides a provisional analysis of the costs of options b, c, d and e;
            • Section 4 provides a provisional analysis of the benefits of options b, c, d and e;
            • Section 5 contains a summary of their potential costs and benefits.
The following section highlights the groups that would be affected by the proposals. These include:


                                                                                                                  4
                    •     Aspiring drivers
                    •     Approved Driving Instructors
                    •     Driving Standards Agency
                    •     Education providers
                    •     Others

These groups are affected in different ways and to different extents and not all groups are affected by the
different options as indicated schematically in Table 2.1.
          Table 1.1 Overview of groups affected
                                      Learner        Newly              Driving            Driving           Education         Others (e.g.
                                      Drivers        Qualified          Instructors        Standards         Providers         police, other
                                                     Drivers                               Agency                              road users)

          Option b – changes
          to driver training and            √                ×                  √                 √                 √                  √
          testing

          Option c –
          restrictions on learner           √                ×                  √                 ×                 ×                  √
          drivers

          Option d –
          restrictions on newly             ×                √                  ×                 ×                 ×                  √
          qualified drivers

          Option e – changes
          to driver training and
          testing and                       √                √                  √                 √                 √                  √
          restrictions on newly
          qualified drivers




The information provided relates to Great Britain only as Northern Ireland is not covered by the scope of these
proposals.


1.6     Aspiring drivers

The changes to the training and testing system will impact upon people who are seeking to learn to drive; and,
if options d and e are adopted, those who had recently passed their driving test. In 2006, just over 950,000
provisional licences were issued; over 50% of these were to individuals aged 17 (or 16 – in readiness for their
being allowed to drive at 17). In 2006, there were estimated to be 9.9 million active provisional licences in
existence, 4 although the number of provisional licences which are held does not indicate how many people are
actively learning to drive at the current time.

For the purpose of this study it is assumed on average there are 750,000 people actively learning to drive in
each year. 5 In reality it is clearly a simplification to assume a constant number of people actively learning to
drive in each year. Some of the potential changes in the system – and the associated changes in pass rates
and time taken to learn to drive – may have complex effects on the 'inflows' of new learners, the 'outflows' of


4
  DVLA Database of Provisional Licences (2006)
5
  DSA RSIS database (2006/07) – Average number of newly qualified drivers is around 750,000. DSA figures for the number of car driver licences
issued for the last three years: 2006/07 - 776,156, 2005/06 - 780,907, 2004/05 - 707,018. It is assumed that the annual number of newly qualified
drivers is approximately equal to the annual number of active learners.



                                                                                                                                                    5
those who abandon the effort to learn to drive and the annual stock of continuing learners. Development of a
stock-flow model to take account of all these effects is beyond the scope of this initial study. Information
relating to timing is still subject to further discussion and finalisation within Government.


1.7     Approved Driving Instructors

Registration with DSA as an approved driving instructor (ADI) is a legal requirement for those offering paid
driving lessons; non-registered individuals may only accompany learners; they may not offer paid lessons.
DSA records indicate that there are currently 42,144 ADIs. 6 The number of ADIs has been increasing in recent
years and research in 2000 estimated that around 2,500 new ADIs qualify each year, with 2,000 leaving the
industry every year. 7 A survey of ADIs (undertaken in 1998) found that the vast majority (84%) described
themselves as self-employed, with three-quarters identifying driving instruction as their main employment.


1.8     The Driving Standards Agency (DSA)

The DSA is responsible for administration and delivery of the driver testing function. There are currently 1,984
driving examiners 8 employed by the DSA. However, more examiners are required to deal with current levels of
demand and many examiners are approaching retirement age so it is recognised that steady recruitment is
needed if demand for tests requires DSA to maintain current numbers of examiners.

The DSA conducts practical driving tests from a national network of 360 car driving test centres. 9 The DSA is
planning to establish a network of Multi-purpose Test Centres located in areas with easy access to the
regional road network and within a short travelling time of a variety of road conditions, allowing the driving
examiners to assess each candidate across the full range of driving skills. There are 158 theory test centres
across Great Britain. Theory test delivery has been contracted out to a private company but the DSA retains
the responsibility for the control of the theory test and maintenance of the question bank. 10


1.9     Education providers

Only option b places an emphasis on shaping and promoting better driving attitudes through voluntary pre-
driver education courses and a structured learning to drive syllabus. As yet, the scope of provision in relation
to the proposed new access and learning centres has not been decided although it is proposed that pre-driver
education will be carried out predominantly at schools and other educational establishments, as part of the
extended school day. 11 Potentially the number of delivery outlets could range from around 75 – one for each
area covered by a Multi-purpose Test Centre - to 5,000 the number of educational establishments dealing with
the 16-plus age group. Dependent on findings from the consultation, there is scope for delivery of pre-driver
education online. It is expected the training would be largely similar to the pre-driver qualification provided in
schools. Some elements will necessarily be different, for example peer group activities would need to be
revised, but the aim will be to cover the same topics by different means.




6
  DSA ADI National Register (October 2007).
7
  Silcock et al (2001) Raising the Standards of ADIs: review of requirements for training and qualification as an Approved Driving Instructor. DfT: London
8
  DSA provided headcount figures of driving examiners (2008)
9
  Available at http://www.dsa.gov.uk/AtoZDTCInfo.asp?letter=A&CAT=-1&s= [accessed 25th January 2008]
10
   The bank of theory test questions has been published since the inception of the test in 1996. It includes all questions that may be asked in the theory
test. At present it is publicly accessible.
11
   Information supplied by DSA indicates that the proposals would be delivered through existing channels but is still subject to further discussion.



                                                                                                                                                         6
The public consultation will explore further the cost and benefits of the potential options for delivery of access
and learning centres and will review in greater detail the likely effectiveness of different approaches.


1.10 Others

It is anticipated that the Driver and Vehicle Licensing Agency (DVLA) will maintain its existing role in issuing
provisional licences, of which almost 1 million new licences are currently issued each year. It is expected that
the improved driving standards and attitudes arising from all the options would result in reduced demands on
police/legal system resources in relation to accidents and road traffic offences. However, any new restrictions
on new drivers, in particular, from option d would require police resources to enforce.

Other road users as a whole would be expected to benefit from better drivers and improved road safety with a
reduction in accidents and a potential reduction in insurance premia.




                                                                                                                     7
2 Provisional analysis of costs

The costs of improved training and testing will depend heavily on how extensive any change is. The figures
provided here are estimates of what it might cost to implement a version of the new system. However, much
will depend on decisions taken after the consultation paper, both on matters of policy such as the content of
the test, and on operational matters, such as the speed of implementation.

In all the options there will be instances where some people do not comply with the regulation, and thus
avoiding the costs of compliance. However, as a result this group will have to deal with the risks of non-
compliance, such as the risk of arrest, or a substantial fine, and consequently may need to modify driving
behaviour. These can be described as non-compliance costs. If non-compliance takes place, it appears that
on average compliance costs are likely to be higher than non-compliance costs. As such, the figures given
here should be considered as an upper-bound to the costs of regulation.


2.1        Option a – Do nothing

No additional costs would be expected under the do nothing scenario as the current training and testing
system would be maintained. A summary of the methodology and estimate of the number of casualties
involving newly qualified drivers and their passengers is set out in Annex A.


2.2        Option b – Changes to car driver training and testing

2.2.1     Voluntary pre-driver education
Pre-driver education is voluntary and is expected to be delivered in schools/learning centres. It is assumed
that there would be no substantial direct costs for learners as most will be learning as part of the extended
school day, with course materials, etc provided. For learners above school age the learning will be varied and
will continue to be delivered through books such as the Highway Code. Either route would, of course, involve
some opportunity costs, if only in terms of forgone leisure activities.

2.2.1.1 Pre-driver qualification
The costs associated with the pre-driver foundation certificate are not certain, the DSA have estimated that it
will entail a one-off cost between £4.8m and £6m to develop the pre-driver education programme.

The annual operating costs of running the pre-driver education programme will be dependent on the
participation rate. Initial feedback from the consultation workshops with young drivers has indicated a healthy
interest in such a course. As no information is available on the likely participation rate by young people, the
scenario below assumes that, of the annual cohort of 700,000 14 year olds, 12 a quarter will undertake the
voluntary training at a cost of £200 per pupil. 13

          Estimated Number of Pupils undertaking Pre-driver Education (175,000) X Cost of Course per Participant (£200) = Cost of
          operating Pre-driver Qualification Programme per annum (£35m)




12
  For this scenario 14 year old school children will be targeted for pre-driver education
13
  DSA indicated that courses for remedial drivers cost £100 per lesson per class. It is estimated that pre-driver qualification will be taught in classes of
around 20 pupils, and that it will take a maximum of around 40 hours to teach, in line with equivalent qualifications. £100 per class / 20 participants x 40
hours = £200 per pupil.



                                                                                                                                                           8
It is not expected that the pre-driver training will have a significant effect on the number of lessons undertaken
by a learner, although it may assist in the improving the pass rate of the theoretical components of the car
driving test for these learners.

2.2.2        Voluntary learning to drive
2.2.2.1 ADI training for the new system
The development of a new car driver training and testing regime assumes that the retraining of ADIs will be
required to assist learners. It is not possible to produce an accurate estimate of the cost of retraining
instructors until there is greater certainty about the form of the new system. Much will depend on which new
elements get adopted into the new assessment framework.

The DSA expects that most of the training that helps driving instructors adapt would form part of the
Continuing Professional Development (CPD) programme, together with modified training for new applicants.
While it is not certain that CPD will necessarily form the basis of training for the new system, were another
route chosen, it would probably entail separate costs of a similar magnitude.

Exact costs are not currently available. While CPD costs are still being developed in consultation with the
driver training industry, an initial estimate by the DSA proposes that the one off cost of developing and quality
assuring CPD courses, together with the costs of marketing to instructors and giving them an incentive to take
up training, would cost between £4.3m and £10.7m. 14

2.2.2.2 Developing an attitude indicator
The attitude indicator is a voluntary online assessment tool enabling learners to examine their attitudes to a
range of scenarios and encourage better driving habits. Based on experience with other assessments, DSA
estimates that the development of the attitude indicator will cost around £1m. This reflects the costs of
employing behavioural experts to produce the test itself, as well as the costs of creating the supporting
computer infrastructure. As there is less need to produce an extensive question bank, to update it, or to keep it
secure, DSA expects developing an attitude indicator to be cheaper than changes to the hazard perception
test and theory test. The form of the attitude indicator needs to be decided before ongoing costs can be
estimated.

2.2.2.3 Peer group work
Learner drivers would be encouraged to attend voluntary peer group sessions to discuss and develop better
and safer driving attitudes. Based on experience from remedial courses, if a peer group session were to be
included, the cost of providing the session would be estimated at around £10 per participant. It is assumed that
learner drivers would only attend 1 session. As it is voluntary the cost of providing this intervention would be
dependent on the take up rate by learner drivers. For the scenario below, if a quarter of learner drivers were to
be involved in peer group work then it would cost £2m per annum.

             Estimated Number of Learners attending Peer Group Sessions (187,500) 15 X Cost per Participant (£10) = Cost of Providing
             Peer Group Course per annum (£1.9m)

2.2.3        Mandatory standards and testing
For option b the elements relating to standards and testing will be mandatory for all learner drivers.




14
     This takes into account that ADIs attending training will face opportunity costs in terms of lost earnings
15
     Calculated by the annual number of learner drivers 750,000 multiplied by the take up rate in this scenario of 25%



                                                                                                                                        9
2.2.3.1 Competence framework
The costs for developing the Competence Framework for assessing learner drivers have already been
incurred.

2.2.3.2 Learner workbook
Workbooks will be provided to learner drivers to support them in working through the driving task and enable
learners to monitor their progress. Workbooks will be produced on a large print run, which will limit their total
cost. The most expensive production options would be expected to cost a maximum of £5 per copy. Were the
DSA to issue a copy of this workbook to every learner, the total cost would be £3.75m per annum.

          Number of Learners (750,000) X Cost of Workbook (£5) = Total Cost of Workbook per annum (£3.75m)

2.2.3.3 New theory test items
Based on previous experience with producing case studies used in the new theory tests for lorries, buses and
coaches, it appears from DSA that developing case studies for the car theory test would cost approximately
£1.8m. Ongoing costs cannot be estimated until further scoping work has been completed.

2.2.3.4 New hazard perception test
Based on their experience of developing the original Hazard Perception Test, DSA has estimated that
development of a new test using 3D animation would cost approximately £2m.

2.2.3.5 Developing a new practical test
The precise costs are not available at this stage. DSA have estimated that the total cost of implementing the
new practical test is between £2.1m and £5.7m. 16

2.2.3.6 Implementing a new practical test
A new practical test would require a one off cost of retraining approximately 2000 car driving examiners less
an average of the 150 who leave each year due to retirement/old age. This could involve between one and
three weeks with a cost of between £1.9m and £5.6m, not including lost productivity. 17

          Number of Driving Examiners less natural turnover (2000-150) x Cost of Retraining (£1,000-£3,000) = One-off Cost of
          Updating Existing Examiners (£1.9m-£5.6m)

A new practical test could also involve significant changes for DSA back office staff, technical support, booking
systems and corporate functionality. These costs are impossible to quantify at such an early stage. However,
the DSA have estimated that the one off cost of reorganisation and retraining could be between £12.2m and
£21.3m.

2.2.3.7 Increased length of practical test
This is still subject to further discussions is the introduction of a longer practical test in order to accommodate
new elements, for example, independent driving. If this were to happen, an examiner might be able to only
take six tests a day as opposed to the current seven. This would necessitate the recruitment of further
examiners. It would also entail an increase in the test fee, as DSA is a self-financing Trading Fund. Assuming
that costs increase in a linear fashion, a move from a seven test schedule to a six test schedule would
increase the cost of administering a practical test by 16.7%. 18 This would raise the cost of an individual car test
by approximately £8 to £56.60.


16
   Cost estimates provided by DSA
17
   Cost estimates provided by DSA
18
   Percentage calculated by dividing the reduction in tests per day (7-6) by the forecast number of tests per day (6)



                                                                                                                                10
          Cost of Existing Practical Test (£48.50) X Predicted Percentage Increase of Longer Practical Test (1+16.7%) = Estimated
          Cost of Longer Practical Test (£56.60)

Overall DSA costs would be expected to rise by a similar percentage. In 2006/07 the DSA received £90.5m
from car testing fees 19 and the introduction of a longer test would lead DSA costs to rise by a similar
percentage of an additional £15m per annum.

          Existing DSA Cost of Car Testing (£90.5m) X Forecast Percentage Increase (16.7%) = Additional Cost of Car Testing per
          annum (£15.1m)



2.3        Option c – Restrictions on learner drivers

2.3.1     Minimum learning period of 12 months
The proposed minimum learning period of 12 months and increasing the full driving age to 18 years would
involve welfare costs to learner drivers. Research from the Cohort II study reveals that the median time
between starting to learn to drive and passing taken by respondents is 10 months, though this varies by age
and sex. Older age groups are likely to have a progressively longer learning period as reflected in their larger
number of practical test attempts and greater number of hours spent practising with either professional
instruction or with friends/relations. Information from Cohort II 20 also indicates a gender differential with most
males passing their practical test in 8 months in contrast to females who pass in 11 months. Therefore, a
minimum learning period of 12 months will disproportionately affect young drivers (and in particular males)
who annually represent the largest group of learner drivers. 21

Given an average time taken to obtain a licence of 10 months, the effect of a minimum learning period of 12
months is the removal of the value newly qualified drivers attach to a licence for the average of the two months
they otherwise would have been driving 22 . At present, the extent to which new drivers attach a value to a
licence is unclear so a range has been constructed on the assumptions below. The scenario assumes that the
annual number of new licences is 750,000 23 and the annual value of a full driving licence for new learner
drivers is between £800 and £1,500. 24 This would involve an annual welfare cost of £100m-£188m.

          Loss of Value of Licence for Two Months (£133-£250) 25 X Annual Number of Newly Qualified Drivers (750,000) = Estimated
          Welfare Cost of Proposed Minimum Learning Period (£100m-£188m) per annum

2.3.2     Full licence driving age of 18 years old
As with the minimum learning period, increasing the minimum age of a full driving licence to 18 years would
involve welfare costs to newly qualified drivers. On average, under-18 years old obtain their licence within 9
months and the effect of increasing the minimum age for acquiring a full driving licence to 18 years is the loss

19
   Assumption that the income from fees covers all costs associated with car driver testing including overheads, research and development, etc.
20
   Wells P, Tong S, Sexton B, Grayson G, and Jones E (2008) Cohort II: a study of learner and new drivers. Road Safety Research Report 81. DfT.
21
   DVLA Database of Newly Qualified Drivers (2007). 55% of provisional licences issued in 2006 were to 16 and 17 year olds. 18 year olds accounted
for only 8% of provisional licences.
22
   These figures are, of course, no more than approximations since the precise impact depends upon the distributions involved – which are unknown –
not just the mean values.
23
   DVLA Database of Newly Qualified Drivers (2007). Average number of new licences issued 2004-2006.
24
   The lower figure of £800 is taken from a study on Norwegian drivers’ willingness to pay for not losing their driving licence for 12 months (after
exchange rate conversion) - from Jorgensen, F. & Dargay, J., (March 2006) ‘Inferring Price Elasticities of Car Use and Moral Costs of Driving without a
Licence,’ Oxford University. The higher figure of £1,500 is based on a DSA estimate of the average cost of learning to drive – cost of provisional licence,
test fees, lessons with a professional instructor, insurance, petrol, etc. which approximately totals £1,500. For this assessment the higher figure is used
as a recognised benefit of the annual value though this investment is expected to generate returns over many years so may well exaggerate the annual
value.
25
   The value of licence for two months. This is calculated by the value of the licence (£800-£1500) divided by twelve months multiplied by the two
months the average driver would have been on the road if not prevented by the minimum learning period.



                                                                                                                                                        11
of the value of the licence for the three months those involved would be unable to drive alone 26 . Again, the
extent to which under-18 year olds attach a value to a licence is unclear. The scenario below assumes that
annual value of a driving licence for under-18 year olds is between £800 and £1,500 (applying the same
assumptions as outlined in the minimum learning period) and that the annual number of under-18 year old
obtaining a full licence is 412,500 27 .

          Loss of Value of Licence (£200-£375) [3] X Annual Number of Newly Qualified Drivers under 18 (412,500) = Estimated
          Welfare Cost of Proposed Minimum Driving Age (£83m-£155m) per annum

A minimum driving age of 18 years would involve a welfare cost of £112m-£211m to individuals under 18 who
currently obtain a full driving licence. However, this total cost is not additive to the total cost of the minimum
learning period as one is the subset of the other. The welfare cost of operating a minimum learning period of
12 months along with a full driving age of 18 years old is £127m-£239m.

                [Loss of Value of Licence (£200-£375) [4] X Annual Number of Newly Qualified Drivers under 18 (412,500)] + [Loss of
                Value of Licence (£133-£250) [5] X Annual Number of Newly Qualified Drivers above 18(337,500)] = Estimated Welfare
                Cost of Proposed Minimum Driving Age and Learning Period (£127m-£239m) per annum

This scenario assumes that 337,500 individuals – 45 per cent of the total annual number of learner drivers –
are above 18 years old and would be affected by only a minimum learning period of 12 months which
diminishes the value of a full licence by 2 months. The remaining learner drivers would be affected by the
introduction of both a minimum learning period and a minimum full driving age of 18 years old, which would
diminish the value of a full licence by 3 months. In addition, the proposal of a minimum driving age of 18 would
prevent some 17 year olds from undertaking work activities or restrict the job opportunities available to them.

2.3.3     Minimum number of lessons
The proposal of a minimum number of lessons will be borne primarily by those who have no or minimal contact
with an ADI for their driver training. The cost of the proposed minimum number of lessons with ADIs is
dependent on the number required, which, in turn, is dependent on the identified purpose for ensuring learner
drivers receive professional instruction. Further consultation is required to establish the likely minimum
requirement though discussions from DSA and DfT have indicated that two scenarios should be considered:

          • 10 lessons – to help learners to structure and complement their learning.
          • 36 lessons – the current median number of lessons for people to pass their test 28 .

In practice, the lower range of proposed minimum number of lessons, seems unlikely to affect those learner
drivers who currently use ADIs. Responses from Cohort II 29 revealed that on average those sitting the practical
test will have had 52 lessons with a professional instructor. A high number of minimum lessons will result in
additional costs being borne by a substantially larger number of learner drivers, generating a much larger
additional cost. The scenario of 36 minimum lessons could by definition affect around half of learner drivers.

The scenarios below apply indicative figures for the percentage of learner drivers requiring additional lessons
in each case.


26
   These figures are, of course, no more than approximations since the precise impact depends upon the distributions involved – which are unknown –
not just the mean values.
27
   DVLA (2007) Approximately 55% of provisional licences are issued to under 18 year olds. Due to the limitations of the data available in relation to the
age profile of people actively learning to drive it is assumed for this study that the breakdown of provisional licences by age groups is indicative of the
profile of learner drivers. In reality this is an oversimplification.
28
   Wells et al (2008)
29
   Cohort II is not representative of all learner drivers but due to the lack of available data is used as indicative of general learner drivers



                                                                                                                                                         12
     (i) 10 lessons

     Number of Learners Assumed to be Affected (750,000 X 4.2%) 30 X Cost of 5 Lessons (£105) 31 = Annual Cost of Minimum of 10
     Lessons (£3.3m)

This would involve an additional cost to learner drivers of around £3m per annum. This figure assumes a small
proportion <5% of learner drivers will require a modest number of additional lessons with ADIs to comply with
the proposed requirement.

     (ii) 36 lessons

     Number of Learners Affected (375,000) 32 X Cost of 12 Lessons (£252) 33 = Annual Cost of Minimum of 36 Lessons (£94.5m)

This scenario of a minimum of 36 lessons would result in an additional cost to learner drivers of £95m per
annum. This estimate assumes that half of learner drivers would require a significant average number of
additional lessons from ADIs.

2.3.4     Reducing the provisional driving age to 16.5 years
Reducing the provisional licensing age to 16 and half years may increase the risk of casualties and damage
only accidents involving young learner drivers. Though the information relating to number of incidents is not
well evidenced, DfT sponsored research 34 estimates that an additional 130 killed or serious injured
casualties 35 per annum could arise as a result of reducing the provisional age. This research provides no
information in relation to slightly injured casualties; however, using the Road Casualties 2006 Annual Report 36
and applying the ratio of slightly injured casualties to killed and seriously injured casualties from Road
Casualties 2006, the total number of casualties can be estimated as shown in the below table.

The number of damage only accidents can be estimated from the number of injury casualties 37 using
assumptions from the Highways Economic Note No1. 38 Although there is no reliable information on these
numbers as they are not comprehensively reported, it is estimated that, if the total number of additional injury
accidents is 1,610, then the number of damage only accidents would be around 23,300, which is calculated by
applying a factor of 14.5 to casualty accidents.

Applying the monetary values attached to the prevention of road accidents as outlined in Highways Economic
Note No1 (adjusted to 2006 prices), 39 the cost of additional casualties and damage only accidents that might

30
   Cohort II found that 4.2% of passers received between 0-10 hours of professional tuition. It is assumed that these learners will all require an average
of 5 additional lessons of ADI instruction.
31
   The figures assumes that on average the learner drivers affected would require 5 more lessons than at present at a cost of £21 per lesson. For this
study, £21 is the suggested cost of a one hour lesson, This is derived as an educated estimate from experienced DSA staff given the varying
geographical locations and levels of skill of ADIs
32
   The calculation assumes again 750,000 learner drivers annually of which half will require additional ADI instruction
33
   The figure assumes that on average the learner drivers affected would require 12 more lessons with ADIs than currently at a cost of £21 per lesson.
34
   Annex B
35
   The baseline of drivers aged between 17-19 results in a total of 3615 KSI casualties; if provisional licensing is introduced it is estimated that KSI
casualties for the up to 19 age group would increase to 3745 (Annex A and B)
36
   Road Casualties GB 2006. Table 6c: car drivers KSI 9305, all severities 115,003, therefore a ratio of 1:11.359 KSI:Slight Injury is adopted
37
   DfT(2007) Highways Economic Note No.1 advises that damage only accidents are to be calculated assuming an average of 17.7 damage only for
every injury accident on built-up roads, 7.8 on non built-up roads and 7.6 on motorways. From the GB Road Casualties 2006 the proportion of accidents
on built-up roads is 68%, non built-up road accidents is 27% and on motorways it is 5%. Multiplying the corresponding proportion of road type accidents
by the suggested rates results in an overall multiplier of 14.5, which can be applied to injury accidents to derive a crude estimate of damage only
accidents.
38
   DfT (2007) Highways Economic Note No.1
39
   DfT (2007) Highways Economic Note No.1 - the average value of casualty type is: killed £1,428,180; seriously injured £160,480; slight £12,370; and
average value of damage only accident is £1,710. These values are 2005 prices, using the recommended inflator of 1.0427, 2006 prices can be derived,
which are rounded to the nearest £10 in line with values in the Highways Economic Note.



                                                                                                                                                       13
occur because of the reduction in the provisional driving age can be estimated. The potential annual cost of
casualties is £61m, of which £42m relates to killed and seriously injured accidents, whilst damage only
accidents might result in an additional cost of £41m per annum as shown in Table 3.1. The value of types of
casualties and accident types is based on 2006 prices.

          Table 2.1 Estimated cost of additional casualties and damage only accidents due to provisional
          licensing age of 16.5 years
           Casulaty / Accident Type            Additional Number of                Value of casualty / accident           Total Value (£000s)
                                               Casualties / Accidents              type (£)

           Killed                                             15                               1,489,160                              22,337

           Seriously Injured                                 115                                167,330                               19,243

           Slight                                           1,480                                12,900                               19,092

           Damage Only                                     23,300                                 1,780                               41,474

           Total Casualty Cost                                                                                                        60,672

           Total Casualty and                                                                                                        102,146
           Accident Cost

          Source: DfT Road Causalities Great Britain 2006/Highways Economic Note No.1 2005


2.4        Option d – Restrictions on newly qualified drivers

2.4.1     Enforcement of restrictions
The proposed restrictions for newly qualified drivers are likely to result in additional costs to enforcement
agencies such as the DVLA, the police and the courts, although the resources involved are impossible to
quantify and mostly represent the opportunity cost of diverting resources from other uses.

2.4.2     Value of restricted licences
The restrictions on licences would involve welfare costs to newly qualified drivers that limit the social benefits
associated with driving. These costs will vary dependent on the length of time the restrictions are in place. The
extent of the welfare cost is uncertain and a range has been estimated based on the scenarios below.

In the absence of real evidence it is assumed that a restricted licence is worth half the value of a full licence
and using the licence values used in section 3.3.1 implies that the value of a licence to new drivers is reduced
by £400-£750 per annum for the period the restrictions remain in force. 40

          Reduction of Value of Licence (£400-£750) X Number of Newly Qualified Drivers (750,000) = Annual Welfare Cost of
          Restrictions on Drivers (£300m-£563m)

Therefore, a proposed restriction period of 1 year would involve an annual welfare cost of £300m-£563m. This
is derived from the number of newly qualified drivers each year -750,000 – multiplied by the decrease in value
attached to the licence. Restriction periods of 2 years and 6 months would result in annual welfare costs of
£600m-£1,126m and £150-£281m respectively based upon analogous assumptions – i.e. double or half the
estimated cost of a one year restriction respectively.


40
  In practice, the value of a restricted licence is dependent on the precise restrictions that are put in place and the circumstances of the user - for
example, a passenger night time curfew on driving may affect the employability of some individuals



                                                                                                                                                          14
2.5     Option e – Changes to driver training and testing and restrictions on newly qualified
      drivers

This is a combination of all the costs outlined in options b and d.




                                                                                                15
3 Provisional analysis of benefits

For each of the 'do more' options the major recurring impact is likely to be the cost savings generated by the
reduction in the likelihood of casualty and damage only accidents by newly qualified drivers and their
passengers from the proposed changes. The data that would be required to identify exactly how many
accidents are caused by newly qualified drivers is not readily available. Casualty statistics collated by the
police include the age of drivers and casualties, but not the date on which their licence was issued. Given the
limitations associated with the STATS19 data source and the difficulty in eliminating double-counting from the
data, DfT have estimated a baseline on the numbers of drivers and their passengers killed and seriously
injured that is applied for this impact assessment. 41 The DfT is hoping to commission further research to
improve the methodology to estimate all casualties in accidents involving newly qualified drivers.

In discussions with DSA and DfT, scenarios have been constructed for the potential savings in casualties and
damage only accident costs by newly qualified drivers. This approach reflects the limited evidence or relevant
research on the effects of comparable interventions, as well as the uncertainties around likely levels of
compliance.

An assumption of full compliance by learner or newly qualified drivers to potential new requirements is
unrealistic. It seems reasonable to assume that compliance by learner drivers will be less of an issue than
compliance by newly qualified drivers. Learners will not typically have their own car and many will rely on
parents for access to a vehicle. Parents will be strictest when someone is learning to drive and are unlikely to
tolerate substantial non-compliance. Once drivers have passed their test, and/or have their own, or have
increased access to, a vehicle then non-compliance rates seem likely to be higher.

There is no available evidence to clearly indicate the likely compliance rate for each of the options and the
analysis needs to be based on assumptions. There is a concern that non-compliant drivers are quite likely to
be those who would speed or drive in a more risky way and hence would contribute to a higher proportion of
incidents as compared to those who always obey the law and are the safest drivers. Therefore, it is plausible
for a situation to arise where there is a large degree of compliance amongst the general population but little
compliance amongst those who are most likely to be involved in death or serious injury accidents, limiting the
benefits of potential changes to the driver testing/training regime.

From discussions with DSA, Table 4.1 below sets out assumed non-compliance rates for each of the
options. 42 The application of general non-compliance rates involves two potential biases operating in different
directions: it seems likely - and the DSA figures bear this out – that non-compliance rates are likely to be
highest amongst the groups who present the greatest risks; conversely the restrictions may well have some
effects even on those who do not fully comply with the restrictions.

      Table 3.1 Potential rates of compliance
           Option                                    Possible non-
                                                     compliance rates
           Option b – Changes to driver                        3.5%
           testing and training
           Option c – Restrictions on learner                  15%


41
     See Annex A
42
   Initial discussions with DSA provided estimated non-compliance rates for the general population and separate suggested non-compliance rates for
killed and serious injured accidents. As the assessment includes the effect on damage only and slightly injured accidents as well, the non-compliance
rates for the general population throughout were adopted.



                                                                                                                                                        16
           drivers
           Option d – Restrictions on newly                       50%
           qualified drivers



Option b assumes a similar rate of non-compliance to that seen today as the pre-driver educational training,
attitude assessments are non mandatory whilst there are no additional compulsory components compared
with the current regime. The analysis for option c reflects the assumption that only a relatively modest rise in
non-compliance is likely amongst learners whilst in relation to option d a high degree of non-compliance is
envisaged due to the nature of restrictions envisaged and the practical difficulties of enforcement.

For each option the estimate of cost savings from prevention of casualties and damage only accidents is
derived from the following calculations:

          Baseline X (1-Non-compliance Rate) X Assumed Casualty or Damage Only Accident Reduction Rate Amongst Compliant
          Drivers = Annual Number of Casualties or Damage Only Accidents Prevented

          Annual Number of Casualties or Damage Only Accidents Prevented X Value of Prevention per Casualty or Damage Only
          Accident = Annual Cost Saving from Casualty or Damage Only Accident Prevention.

As already stated, the baseline estimates casualties for car drivers and their passengers only, as such the
reduction in casualties might be an underestimate of potential benefits. However, factors such as
displacement and attribution of blame for accidents may overrate the quantified benefits. Furthermore, the
estimated savings are likely to be overestimated to the extent that non-compliance rates are higher amongst
the riskiest drivers. Conversely, the likelihood is that the restrictions will have at least some effects on the
behaviour of the 'non-compliant.'


3.1        Option a – Do nothing

No additional benefits would be expected under the do nothing scenario as the current training and testing
system would be maintained. For the purpose of this assessment a baseline in terms of casualties, which
represents the “Do Nothing” Reference Case, has been constructed based on information supplied by the
DVLA and Road Casualties Great Britain 2006. DfT sponsored research has estimated that approximately 337
car drivers and passengers are killed per annum in accidents involving drivers with less than two years
experience and 2414 are seriously injured (see Annex A).

Using the information from the Road Casualties 2006 Great Britain and the Highways Economic Note no.1
2005 the number of slightly injured and damage only accidents can be estimated, which are then rounded
given the imprecise nature of the estimate.

          • To estimate the number of slightly injured a ratio of 1:11.359 is applied to the number of killed and
            seriously injured casualties. 43
          • To estimate the number of damage only accidents a factor of 14.5 is applied to the total number of
            casualties (killed, seriously injured and slightly injured). 44



43
   This is the same ration as applied in section 3.3.4 to estimate the number of slight injuries as a result of reducing the provisional licensing age derived
from Road Casualties Great Britain 2006
44
   This is the same factor as applied in section 3.3.4 to estimate the number of damage only accidents as a result of reducing the provisional licensing
age derived from Highways Economics Note No.1 2005 Valuation of the Benefits of Prevention of Road Accidents and Casualties



                                                                                                                                                           17
Table 4.2 establishes a baseline estimate of casualties and damage only accidents involving newly qualified
drivers and their passengers. Assigning monetary values based upon Highways Economics Note No.1 45 , the
cost of casualties and damage only accidents by newly qualified drivers is estimated to be around £2.2 billion
per annum.


     Table 3.2 Average value of prevention per casualty and per accident GB 2006 prices (estimated for
     newly qualified drivers and their passengers)
          Casualty / Accident Type          Value of casualty or damage          Annual number of               Total Value (£000,000s)
                                            only accident type (£)               casualties / accidents

          Killed                                       1,489,160                             337                               502

          Seriously Injured                             167,330                             2,414                              404

          Slight                                         12,900                            31,250                              403

          Damage only                                     1,780                            493,000                             877

         Source: Highways Economic Note No.1 2005


3.2       Option b – Changes to driver training and testing

3.2.1    Faster learning
A possible scenario for learner drivers who voluntarily undertake a more structured approach to learning, for
example, undertaking pre-driver education, attending peer group sessions is that improved training could lead
to learners being prepared for the test in fewer lessons. Experience from across education shows that modular
approaches to learning are more efficient ways to communicate subject matter.

Car learner drivers are not expected to take the structured approach to a specific speed. Currently, the
average number of lessons taken by a learner driver sitting a practical test is 52. Assuming the more
structured approach is similar to the Dutch ‘stepwise’ programme, which consists of 39 elements and that the
structured approach worked at a rate of one element per lesson, this would reduce the average number of
lessons taken by participants by 13. The scenario below assumes that a quarter of learner drivers would adopt
this method of learning resulting in potential savings of £51m per annum to learners.

         Number of Lessons Saved (13) X Cost per Lesson (£21) X Number of Participants (187,500) 46 = Cost Saving to Learner
         Drivers who adopt Structured Approach (£51.2m) per annum

3.2.2    Higher pass rate for structured learning
Participants who follow the structured approach should be more likely to pass their practical test. The Dutch
‘stepwise’ programme has a pass rate of 60-65%, 15-20 percentage points above the regular pass rate. If
such an improvement was found among participants in the structured approach, this would mean 118,000
fewer unsuccessful tests. 47 This scenario assumes that the pass rate for those adopting a structured approach
rises to 60% from the current rate of 43%.

         Number of Fewer Unsuccessful Tests (118,125) X Cost of Practical Test (£48.50) = Cost Saving of Practical Test Fees for
         Learners using Structured Learning (£5.7m) per annum.

45
   Values are adjusted to 2006 prices as in section 3.3.4
46
   Assumption that a quarter of annual learner drivers (750,000) will undertake structured learning approach
47
   Calculated based upon the difference between test numbers of 60% and the current pass rate of 43% of learners using structured approach [(187500
x 1.67) = 313,125 less (187500 x 2.3) = 431,250] = 118,125 fewer unsuccessful tests



                                                                                                                                                 18
Participants would thereby save around £6m per annum in test fees, plus £15m per annum for further lessons
they would have had to take before their retest.

          Number of Fewer Unsuccessful Test (118,125) X Number of Lessons between Retakes 48 (6) X Cost per Lesson (£21) = Cost
          Saving of Lesson Fees for Learners using Structured Learning (£14.9m) per annum.

The improvement in pass rates may also involve welfare benefits if it enables learners to acquire their licence
more quickly, although this effect cannot realistically be quantified.

3.2.3     Improved pass rate for non-structured learning
According to the DSA, a possible impact is that learners who do not take the structured approach may still find
it easier to pass their practical test, because of the information contained in the syllabus and the workbook. If
this were to amount to a five percentage point increase 49 in the pass rate among those not taking the
structured route, that would amount to 124,000 fewer failed tests 50 entailing a saving of £6m per annum. This
scenario assumes that three quarters of annual learner drivers (750,000 X 0.75 = 562,500) will adopt a non
structured approach to learning and work through the syllabus and workbook thereby increasing the pass rate
for the practical test for these learners from the existing level of 43% to 48%.

          Number of Fewer Unsuccessful Tests (123,750) X Cost of Practical Test (£48.50) = Cost Saving of Practical Test Fees for
          Learners using Non Structured Approach (£6m) per annum.

Participants would thereby save £6m in test fees, plus £16m of any further lessons they would have had to
taken before their retest per annum.

          Number of Fewer Unsuccessful Test (123,750) X Number of Lessons between Retakes 51 (6) X Cost per Lesson (£21) = Cost
          Saving of Lesson Fees for Learners using Structured Learning (£15.6m) per annum.

It is assumed that the non structured approach would not reduce the current average number of lessons by
learners for sitting the practical test. Again, there may be welfare benefits if learners acquire licences – on
average – within a shorter period.

3.2.4     Decoupling the hazard perception test from the theory test
A more modular approach to testing involving decoupling the hazard perception test from the current theory
test will enable learner drivers to only have to take the tests they fail, resulting in fewer whole tests. Although
the cost of the theory test and hazard perception will increase as separate entities 52 it still has the potential to
deliver cost savings to learner drivers.

          [Number of Learners (750,000) X Cost of Existing Theory Test Fee (£28.50) X Average Number of Attempts Passing the
          Overall Theory Test (1.59) 53 ] – [Number of Learners (750,000) X Cost of Individual Hazard Perception Test Fee (£15) X
          Average Number of Attempts Passing the Hazard Perception Test (1.16) 54 ] – [Number of Learners (750,000) X Cost of




48
   There is no information available about the number of lessons taken between tests, so an estimate of six lessons is used.
49
   No evidence available so for this scenario DSA has assumed that improved syllabus and workbook would increase pass rate from 43% to 48% for
practical test
50
   Calculated based upon the difference between test numbers of 48% and the current pass rate of 43% of learners using non structured approach
[(562500 x 2.08) = 1,170,000 less (562500 x 2.3) = 1,293,750] = 123,750 fewer unsuccessful tests
51
   There is no information available about the number of lessons taken between tests, so an estimate of six lessons is used.
52
   Current cost of Theory Test is £28.50. Estimated individual test fee for hazard perception test and theory test is £15 each though it is still too early for
the DSA to provide realistic costs for decoupling the theory test
53
   The existing overall theory pass rate of 63% is the equivalent to the average learner requiring 1.59 attempts to pass
54
   The existing hazard perception pass rate of 86% is the equivalent of 1.16 attempts to pass



                                                                                                                                                             19
          Individual Theory Test Fee (£15) X Average Number of Attempts Passing Individual Theory Element (1.37) 55 ] = Cost Savings
          to Learner Drivers of Decoupling the Existing Theory Test Elements (£5.5m) per annum.

Assuming that the existing pass rates for each element of the theory test remain unchanged decoupling the
theory test would result in a reduction in the overall number of attempts required by learner drivers to proceed
to the next stage. It is estimated by the savings in test fees from the reduction of attempts that this will result in
savings to the learner driver or around £6m per annum. Against this, there are potential additional access
costs from the need for candidates to make additional trips to test centres; however, participants would not be
discouraged from taking the tests together.

3.2.5     Improved feedback after tests
The primary objective of providing increased feedback to learners after tests is to provide them with a better
understanding of what they must do in order to reach test standard. A potential impact is that it will allow
driving instructors to focus their training better and to prepare learners for their next test in fewer lessons.
Obviously, the reduction in lessons is dependent on the quality of the feedback received.

Assuming that a learner currently takes six lessons between tests and were only to require between a third
and a half as many lessons before their next test, 56 they would save £42-£63 and learner drivers collectively
would save £46m-£59m per annum.

          [Reduction in Number of Retests (265,000) X Practical Test Fee (£48.50)] + [Number of Retests (735,000) X Savings in Cost
          of Lessons to Prepare for Retest (£42-£63)] = Cost Savings to Learner Drivers of Increased Feedback (£46.3-£59.2m) per
          annum

The number of retests is calculated by estimating the total number of tests expected under the new training
and testing regime in option b less the number of successful tests. This assumes that 562,500 learner drivers
using the non-structured learning approach will have a pass rate of 48%, the equivalent of around 1,170,000
tests per annum whilst the 187,500 learners using the structured learning will have a pass rate of 60%, the
equivalent of 313,000 tests per annum. In total, this is almost 1,500,000 car driving tests conducted per annum
- of which the number of failed tests is around 735,000 – the non structured approach generates around
610,000 failed tests and the structured approach around 125,000. These figures are calculated by multiplying
the number of total tests by the failure rate for the respective method of learning. 57

The reduction in the number of retests represents the difference in the number of retests under the current
system and the number of retests as a result of option b. The number of retests under the current regime is
around 1,000,000. This is estimated as the total number of practical tests – over 1,800,000 in 2006/07 58 -
multiplied by the failure rate of 57%. Again, there may be welfare benefits if learners acquire licences – on
average – within a shorter period.

3.2.6     Reduction in the number of road casualties and damage only accidents
As indicated, research suggests that young people pick up attitudes to driving and road safety long before they
reach the minimum age to hold a provisional licence, for example, research conducted by Reading University




55
   The existing theory pass rate of 73% is the equivalent of 1.37 attempts to pass
56
   This scenario assumes that quality feedback can result in a reduction of additional lessons between tests
57
   Failure rate for structured learning is 40%, calculated by 1 minus estimated success rate (60%). Failure rate for non structured learning is 52%,
calculated by 1 minus estimated success rate (48%)
58
   Available at http://www.dsa.gov.uk/Category.asp?cat=1 [accessed 1st February 2008]



                                                                                                                                                       20
for the AA Foundation, showed that many young people of secondary school age, and males in particular,
were showing signs of developing undesirable attitudes to driving. 59

Factors associated with the attitudes, goals and motivations of drivers appear to be important influences on
driver behaviour but are not well addressed in the current training regime. There is a general lack of evidence
to demonstrate how far increased driver training or behaviour modification would improve safety. 60 However
education processes that address these deficiencies may well prove to be effective in improving safety - for
example, the use of focus groups to reduce the social appeal of risk taking behaviour.

3.2.6.1 Compliance issues
As indicated, it is assumed that this option would see a similar level of non-compliance as currently of 3.5% as
previous changes to the driving test have not led to significant increases in non-compliance.

3.2.6.2 Potential reductions in road casualties and damage only accidents involving newly qualified drivers
The impact of this option in relation to reduction of road casualties and damage only accidents involving newly
qualified drivers and their passengers is very difficult to quantify given the lack of available evidence.
Discussions with the DSA indicate that inexperience at independent driving is responsible for 20% of all
accidents in the first two years after the test. 61 By introducing a section of independent driving into the final on-
road assessment, as well as improving training to deal with night driving, it is anticipated that some of the
difficulties of independent driving will be reduced, particularly in the period just after the test when newly
qualified drivers have the highest accident rate.

The casualty prevention rate for option b is shown in the Table 4.3. This is calculated by estimating the
casualty prevention rate for each of the elements within option b. The calculation assumes that a quarter of
learner drivers would take up the voluntary aspects of this option.

          • The modular car training assessment applied in Denmark has been associated with a casualty
            saving of 7% in the first year of driving. 62 Discussions with DSA suggest it is reasonable to assume
            that a similar level of casualty saving (5%-9%) could be realised and potentially more given that the
            structured learning proposed in option b goes further than the Danish model.
          • In the absence of available evidence, it is assumed that pre-driver education could have a safety
            benefit of approximately two-thirds the value realised by the above modular driver training (3%-6%).
          • In terms of improved testing arrangements, the potential safety benefits will be much greater
            amongst those who take the traditional route of learning (10%-20%) but with a limited effect on
            those following a structured learning approach (5%-10%).

          Table 3.3 Calculation of estimated casualty saving rate for option b for newly qualified drivers
          and their passengers
                                                   Structured Take Up              Non Structured Take Up               Combined Casualty
                                                  Rate (25%) X Estimated           Rate (75%) X Estimated                 Savings Rate
                                                  Casualty Savings Rate             Casualty Savings Rate

                 Pre-driver education                   25% X 3%-6%                         75% X 0%




59
   House of Commons Transport Committee (2007) Novice Drivers Seventh Report of Session 2006-07 Volume 1
60
   Baughan C & Simpson H (2002) Graduated driver licensing – a review of some current systems. TRL Report 529
61
   Cohort II found that the accident rate of newly qualified drivers is c0.45 pa and in the next six months is almost half this at 0.24pa. Over the first two
years, an average driver will have 0.53 accidents.
62
   European Commission (2007) Thematic Report: Driver Education, Training & Licensing from ‘Summary and Publication of Best Practices in Road
Safety in the Member States.’



                                                                                                                                                                21
                (voluntary)                           0.75 - 1.5%                           0                         0.75% - 1.5%

                Learning to drive                    25% X 5%-9%                        75% X 0%
                (voluntary)
                                                      1.25 - 2.25%                          0                        1.25% - 2.25%

                Improved testing                    25% X 5%-10%                   75% X 10%-20%
                (mandatory)
                                                      1.25 - 2.5%                       7.5% - 15%                   8.75% - 17.5%

                Overall                                                                                            10.75% - 21.25%




Given that the introduction of the hazard perception training resulted only in a 3-5% reduction in accident
liability among certain groups of new drivers, 63 the above figures may prove optimistic. Therefore, a range of
5-20% for the reduction in accident rates has been applied in the assessment of the benefits of option b. For
each option a broad range is applied for casualty reduction rates reflecting the limited evidence available as
well as the uncertainty and the extent of the effect of options in reducing the number of casualties amongst
newly qualified drivers and their passengers.

DfT indicates that the accident prevention rate for damage only accidents would be higher than the rate for
casualties as option b directly addresses this issue through training and education. Therefore, the accident
prevention rate applied for damage only accidents (10-40%) is doubled the prevention rate for casualties.

3.2.6.3 Quantified benefits of potential reductions in road casualties and damage only accidents
As indicated in Table 4.4, the casualty cost savings relative to the baseline are estimated from assumed
casualty prevention rate – with an adjustment for non-compliance – and the application of DfT guidance costs
from Highways Economics Note No.1 2005 for different accident types.

               Table 3.4 Average value of prevention per casualty and per accident GB 2006 prices (for
               newly qualified drivers and their passengers)
                Casualty / Accident                 Killed               Seriously Injured            Slightly Injured              Damage Only
                Type

                Baseline                              337                       2,414                      31,250                      493,000

                Less non-compliance                   (12)                       (84)                      (1,094)                     (17,255)
                rate of 3.5%
                                                      325                       2,330                      30,156                      475,745

                Casualties prevented                 16-65                     116-466                   1,508-6,031               47,575-191,298
                based on prevention
                rate 5-20% (10-
                20%) 64

                Value of casualty /               24.2-96.9                   19.5-78.0                   19.4-77.8                   84.7-338.7
                accident prevention
                (£000,000s)




Applying this approach, suggests that this option would deliver each year a reduction in killed of 20-70 or
seriously injured casualties of around 120-470 and slightly injured casualties of around 1,510-6,030 resulting in

63
  Wells et al (2008)
64
  The assumption that option b would have more of an effect on damage only accidents as it directly addresses the issue, therefore a higher accident
prevention rate is applied



                                                                                                                                                       22
cost savings of around £63m-£253m per annum. The reduction in damage only accidents would generate
further cost savings of between £85m-£339m per annum. The resultant potential reduction in casualties and
damage only accidents will lead to savings in insurance costs but this will already be reflected in the potential
casualty prevention savings.


3.3            Option c – Restrictions on learner drivers

DfT believes that, for this option to be truly effective, a minimum learning period has to be coupled with action
to ensure that learner drivers are developing experience of a period of sustained driving, either through
satisfactory monitoring arrangements or a sizeable minimum number of lessons with an ADI. However, this
option lacks the guarantee of structured training outlined in option b. A minimum learning period can be used
to enable learning or to protect from risk. The arrangements for this option is not orientated towards more
useful learning but rather to safeguard the wider driving population from drivers who have acquired a short
driving experience before obtaining their full licence.

The analysis of this option assumes that learner drivers are at greatest risk when they start to drive
independently and that this risk gradually reduces to the same level as that of an experienced driver. The best
case scenario is that the minimum learning period imparts experience as effectively as independent driving -
so a learner driver passes their test and has had the same exposure they would currently have after several
months of independent driving. The worst case scenario is that the minimum learning period provides little
effective extra experience and the effect is merely to cut an average of two months from the average driving
'career'. This defines the theoretical range of casualty savings.

3.3.1      Potential reductions in road casualties and damage only accidents involving newly qualified
           drivers
A preliminary study sponsored by DfT provides provisional estimates of casualty savings arising from a
number of scenarios identified by the Road Safety Division. 65 The method and estimates are outlined in Annex
A. 66 The research indicates that a 12 month learning period would see casualties for killed or seriously injured
fall by 11.9%. However, as previously stated, a minimum learning period may merely delay accidents later into
a driver’s career therefore not offsetting the risk.

The same research forecasts that a minimum of 10 or 20 hours of professional instruction would result in less
than a 1% decrease in casualty numbers whilst 40 hours of professional instruction results in an estimated
reduction of 6.6% of accidents.

From the evidence above and employing a cautious approach - to reflect DfT concerns that casualty
reductions arising from a minimum learning period will largely be the result of delaying new drivers entering the
driving population rather than gaining experience – an assumption of a reduction in casualty accidents of only
1-5% from this option appears reasonable if not optimistic. As already stated, a range is applied because of
the limited evidence available and to reflect the uncertainty of the effectiveness of interventions.

3.3.1.1 Compliance issues
As indicated, it is expected that this option would result in a modest level of non-compliance of 15%. This is
based on the assumption that a minimum learning period (option c) would be about as enforceable as the


65
     Annex B
66
  The findings in Annex A have informed the casualty reduction rates applied in options C and D and are based on the midpoint
values identified in the research.



                                                                                                                                23
current provisional licence regulations. Since these are already difficult to enforce it is anticipated there will be
significant opportunities for non-compliance. However, if learners are unable to have their abilities validated by
an examiner, they may be more inclined to judge their readiness themselves. A minimum learning period might
therefore make unlicensed driving seem more acceptable amongst these affected.

3.3.1.2 Quantified benefits of potential reductions in road casualties and damage only accidents
Based upon the compliance and casualty reduction rates suggested above, an annual reduction of 5-15 killed
and 20-100 seriously injured casualties and slight injuries of 270-1,330 could be anticipated. Applying the
same methodology as outlined in relation to option b, this would result in casualty cost savings of £11m-55m
per annum (see Table 4.5).

          Table 3.5 Average value of prevention per casualty and per accident GB 2006 prices (for newly
          qualified drivers and their passengers)
          Casualty / Accident                    Killed              Seriously Injured            Slightly Injured              Damage Only
          Type

          Baseline                                337                        2,414                      31,250                     493,000

          Less non-compliance                     (51)                       (362)                     (4,688)                     (73,950)
          rate of 10%
                                                  286                        2,052                      26,563                     419,050

          Casualty / Accident                     3-14                      21-103                    266-1,328                 4,191-20,953
          prevention rate of 1-5%

          Value of casualty /                   4.3-21.3                   3.4-17.2                    3.4-17.1                    7.4-37.3
          accident prevention
          (£000,000s)




It is assumed in this case that the reduction in damage only accidents would occur at the same rate as
casualty accidents 67 which would generate cost savings of between £7m-£37m per annum.


3.4       Option d – Restrictions on newly qualified drivers

3.4.1     Potential reductions in road casualties and damage only accidents involving newly qualified
          drivers
This option responds to the concern that casualties involving young drivers tend to involve relatively many
injured passengers and that a relatively high proportion of newly qualified drivers have accidents at night. 68
Evidence from other countries indicates that night-time and passenger restrictions can be effective in some
circumstances 69 . However, new laws have usually been accompanied by heightened enforcement, including
by parents controlling access to vehicles, so this evidence may not provide totally accurate guide to the effects
of the restrictions alone.

3.4.2     Compliance issues
This approach assumes that reducing the scope of driving opportunities will result in fewer casualties but it
does not address directly the causes of dangerous/risky driving. This approach is heavily dependent on


67
   It is assumed that damage only accidents are affected less than projected in relation to the proposed better learning and training arrangements;
Cohort II reveals how inexperienced drivers are particularly prone to small damage only accidents.
68
   DfT (2008) Learning to drive: the evidence
69
   Baughan & Simpson (2002) provide examples from North America and New Zealand indicate that accident rates reduced as a result of restrictions



                                                                                                                                                      24
appropriate enforcement arrangements being in place and, as indicated above, it seems reasonable to
assume that the restrictions will involve a relatively high rate of non-compliance. The analysis also assumes
that little displacement will occur - for example, with newly qualified drivers increasing their driving hours to
times outside the curfew. It also ignores the possibility that denying night time driving experience to new
drivers may to an extent merely delay the problems new drivers experience at night.

3.4.3      Quantified benefits of potential reductions in road casualties and damage only accidents
The estimates of casualty savings (Annex A) suggests that the scenario of no passengers between 10 and 20
years old for one year graduated licensing would reduce casualties by 13.1%; a scenario of no night time
driving between 11pm and 5am for the first year would reduce casualties by 17.7%; and the combined effect of
these two scenarios would deliver a reduction rate of 26.3%.This latter figure is likely to be an overestimate, as
it assumes the two factors affect drivers independently, when in practice they tend to occur together.

From the findings above and, again, employing a prudent approach to reflect the limited evidence and
uncertainty, a reduction in casualty accidents range of 10-20% is applied. Factoring in a high non-compliance
rate of 50%, this option would result in an annual reduction of around 20-30 killed and 120-240 seriously
injured casualties and slight injuries of 1,560-3,130. Applying the same methodology as outlined in option b,
this would result in casualty cost savings of £65m-131m per annum (see Table 4.6). Again, however, this will
be an overestimate to the extent that compliance is lower amongst the highest risk drivers.

              Table 3.6 Average value of prevention per casualty and per accident GB 2006 prices (for
              newly qualified drivers and their passengers)
Casualty / Accident Type         Killed          Seriously Injured     Slightly Injured       Damage Only

Baseline                          337                  2,414               31,250                493,000

Less non-compliance rate         (169)                (1,207)             (15,625)              (246,500)
of 50%
                                  169                  1,207               15,625                246,500

Casualty / Accident              17-34               121-241             1,563-3,125          24,650-49,300
prevention rate of 10-20%

Value of casualty /            25.1-50.2             20.2-40.4            20.2-40.3             43.9-87.8
accident prevention
(£000,000s)




It is assumed, as in relation to option d, that the reduction in damage only accidents would occur at the same
rate as casualty accidents resulting in cost savings of between £44m-£88m per annum.


3.5       Option e – Changes to driver training and testing and restrictions on newly qualified
        drivers

The assumption is that this option would deliver the combined benefits identified in options b and d. This
option consists of two distinct elements directed at different stages of a driver’s career. Reflecting the general
assumptions above, it is likely that the driver training programme would have a low non-compliance rate whilst
the restrictions on new drivers might involve a higher degree of non-compliance.




                                                                                                                    25
Assuming this option achieves similar casualty and accident prevention rates as assumed in relation to
options b and d, it would be expected to realise an annual reduction of 20-60 killed and 170-450 seriously
injured casualties and slight injuries of 2,180-5,810.

           Table 3.7 Average Value of prevention per casualty and per accident GB 2006 prices (for
           newly qualified drivers and their passengers)
            Casualty / Accident       Killed         Seriously Injured     Slightly Injured      Damage Only
            Type

            Baseline                   337                 2,414               31,250              493,000

            Less non-compliance        (180)              (1,291)             (16,719)             (263,755)
            rate of 53.5%
                                       157                 1,123               14,531              229,245

            Casualty / Accident       24-63               168-449            2,180-5,813         34,387-91,698
            prevention rate of
            15-40%

            Value of casualty /      35.0-93.4           28.2-75.1            28.1-75.0           61.2-163.2
            accident prevention
            (£000,000s)




Based upon the methodology above this would yield casualty cost savings of £91m-£243m per annum. The
resultant reduction in damage only incidents would yield savings of £61m-£163m per annum. Additionally, this
option would generate cost savings for learners from the potential impacts arising from improved testing and
pass rates as identified in option b of £146m-£158m per annum.

In practice it may be an oversimplification to add the casualty and damage only accident prevention rates from
options b and d and these figures probably represent something of an overestimate of the benefits of the
option.




                                                                                                               26
4 Other impacts

4.1       Competition Assessment

Changes to the driver training and testing regime in option b will help to improve the consistency of delivery
by ADIs, ensuring a better service is provided to aspiring drivers by adopting a more standard and structured
approach to teaching learner drivers. The changes will apply to the whole driver training sector and so would
not generally be expected to adversely impact on competition between ADIs. There may, however, be adverse
effects on the competitive position of those for whom driving instruction is merely a part-time occupation from
the fixed cost in terms of loss productivity of the additional training involved in preparing for the proposed new
training and testing regime.

Wider changes to the driver training sector are only discussed at a theoretical level. Practical proposals for
change will be discussed in a future consultation document.


4.2       Small Firms Impact Test

The driver training industry is dominated by self-employed instructors. Option b will require ADIs to undergo
ongoing training and development. Under option c ADIs would potentially benefit from the requirement of a
minimum number of lessons, though this is expected not to be significant considering the small proportion who
have less than 10 hours of professional instruction. There would clearly be much larger impacts on the
demand for ADI services from requirements for larger numbers of lessons – and at least in the short term –
this could result in higher charges, to the benefit of the industry and the detriment of learners.

To the extent that the changes increase test pass rates and reduce the numbers of lessons to prepare for
retests there may be adverse impacts on the demand for ADIs' services. The net impact on the industry is
clearly uncertain at this stage.

The administration burden to businesses for each option is foreseen to be negligible with little or no changes
from the current situation. The proposed changes are more likely to have a more significant effect upon the
public sector, though these administration costs will largely be transitional.

The impact assessment guidance appears to require that a Small Firm Impact Test is undertaken. This is
beyond the scope of this preliminary study but should be explored as part of the consultation process.


4.3       Environmental impacts (including carbon)

Congestion is a major cost to the economy, and this is expected to rise over time. The Eddington Report on
Britain’s transport future estimates that the increase in congestion between now and 2025 will cost the country
£22bn a year. 70 Poor driving skills and traffic accidents are key causes of congestion, the cost of which is not
included in the financial values attached to accidents.

From option b it is expected that there would be additional environmental benefits from improved driving
standards with reduced emissions and fuel consumption resulting from more efficient driving behaviour. As




70
     DfT and Treasury (2006) The Eddington Transport Study



                                                                                                                 27
part of the pre-driver training, eco-driving learning would be undertaken and DSA sponsored studies 71 indicate
that fuel efficient driving can lead to fuel savings of 5-10%. Smarter driving behaviour allows people to use
roads more efficiently; it can help prevent accidents; and it can mitigate the consequences when they do
occur. This will have some further beneficial impacts on CO2 emissions. Road transport is a significant
contributor to carbon dioxide creation. Vehicle fuel consumption is estimated to have generated 119.9m
tonnes of CO2 in 2005, 21.6% of the national total. 72

Conversely, the higher costs attached to obtaining a licence in options b and c or the cost of a graduated
licence in option d will be of such significance to some learners that they may delay obtaining their driving
licences, limiting their employment and lifestyle opportunities.


4.4     Health impact

Proposals in option b are designed to update the way drivers are trained and tested and as such do not
introduce anything new that would impact directly upon health issues. Better trained drivers will of course
result in a reduction in the number of road traffic collisions, injuries and deaths. Estimates of reductions in
casualties for each of the options are explored in the Section 4 on the benefits.


4.5     Equality impact (including race, disability and gender)

All who qualify to drive must meet the same standard. This means that there is limited scope for differential
impact across the community. It is difficult to identify any way in which sexual orientation, gender re-
assignation or religion could vary the impact of any of these reforms.

People with limited physical mobility have a greater reliance on car travel. We do not propose to change the
current arrangements, in which they can receive a driving licence at age 16. For this reason, there is a high
probability that people with limited mobility would be exempt from any minimum learning period or restrictions
on newly qualified drivers (options c and d).

Revised training and testing (option b) would promote greater equality between the sexes. Currently, women
are less likely than men to pass their driving test. Greater clarity about the learning process and a stronger
focus on safe attitudes should help equalise this imbalance. A minimum learning period (option c) would
impact disproportionately on men; since they tend to qualify more quickly, they would be delayed more by any
restriction. Option d, meanwhile, would be targeted mainly at misbehaving male drivers. Women drivers, who
are generally more risk averse, will receive disproportionately few benefits from graduated licensing, while
paying the full cost.

Options c and d will both have a heavy impact on young people. Three-quarters of those who pass their test
are aged 25 or under. Young people also tend to learn more quickly, meaning they are particularly affected by
a minimum learning period. The vast majority of the costs of such schemes will be paid by those aged 17-27.
This cost will take the form of practical inconvenience and of the discriminatory implication that young people
are inherently unsafe, regardless of their personal behaviour. Balanced against this, the majority of casualty
savings will be among those in this age group. Improved training and testing will have its largest impact
among young people, by making them safer once they have qualified to drive. Again, the majority of casualty

71
   CIECA (2007) CIECA Internal Project on 'Eco-driving' in category B driver training & the driving test Report can be accessed
http://www.ecodrive.org/fileadmin/dam/ecodrive/Downloads/CIECA_Eco-driving_project_final_report_EN.pdf
72
   DEFRA statistics http://www.defra.gov.uk/environment/statistics/globatmos/download/xls/gatb05.xls



                                                                                                                                  28
savings will be among the young. Young people may possibly benefit from more efficient learning
arrangements, leading to lower costs for learning to drive.

As the driving test is a national standard, none of the changes examined in this document would seem to have
a racially discriminatory outcome.


4.6   Human rights

There are no Human Rights impacts to the proposals in option b. A ruling by the Grand Chamber of the
Human Court of Human Rights in June 2007 confirmed that cars have the potential to cause grave injury and
that certain responsibilities therefore come with owning or driving a vehicle.


4.7   Rural proofing

These proposals for driver training and testing apply equally across the country. As such, we do not expect
them to have a significant impact in rural areas compared to other areas over and above the existing
differential. It is acknowledged that the lack of public transport in rural areas means that many people in such
areas may need access to cars for employment and educational purposes. There may be pressure for them
to get through their training and test as fast as possible - that may incur some costs for them. We would be
pleased to hear of any impacts that the proposals might have on those who live in rural areas.



IF YOU HAVE ANY COMMENTS ON THIS IMPACT ASSESSMENT OR ADDITIONAL INFORMATION OR
DATA WHICH COULD HELP IMPROVE OUR ASSESSMENT OF THE COSTS AND BENFITS OF
POLICIES AIMED AT IMPROVING DRIVER TRAINING AND TESTING, WE WOULD LIKE TO HEAR FROM
YOU DURING THIS CONSULTATION. IN PARTICULAR WE ARE KEEN TO FIND OUT MORE
INFORMATION ABOUT:
• Value of licences – what values do learner drivers attach to obtaining and operating a licence?
• Restrictions on licences – to what extent is a licence is devalued by restrictions?
• Total number of casualties involved in accidents with newly qualified drivers or number of casualties
    caused by newly qualified drivers – how can we best estimated these?
• Total number of learner drivers – how can we identify how many people are actively learning in any one
    year?
• Damage only accidents – the current estimate is based on guidance from Highways Economic Note 1 but
    is there a more robust estimate available?
• Compliance rate for each option – is any better evidence available?
• Casualty reduction rate – is there any robust evidence of the effect of each option in reducing casualty
    numbers?




                                                                                                                   29
                                Specific Impacts Test: Checklist

Use the table below to demonstrate how broadly you have considered the potential impacts of your policy
options.

Ensure that the results of any tests that impact on the cost-benefit analysis are contained within the
main evidence base; other results may be annexed.

 Type of testing undertaken                                          Results in            Results
                                                                     Evidence base?        annexed?
 Competition Assessment                                              Yes                   No
 Small Firms Impact Test                                             Yes                   No
 Legal Aid                                                           No                    No
 Sustainable Development                                             No                    No
 Carbon Assessment                                                   Yes                   No
 Other Environment                                                   Yes                   No
 Health Impact Assessment                                            Yes                   No
 Race Equality                                                       Yes                   No
 Disability Equality                                                 Yes                   No
 Gender Equality                                                     Yes                   No
 Human Rights                                                        Yes                   No
 Rural Proofing                                                      Yes                   No




                                                                                                          30
6.0    Summary

The table on the following page summarises the costs and benefits associated with the proposed changes to
the driver training and testing. All the costs and benefits in the table are at 2006 prices.

The summary table (Table 6.1) shows that the largest potential driver of costs relates to the welfare costs from
the potential restrictions on learner and newly qualified drivers. The largest benefits arise from the prevention
of accidents, both casualty and damage only. Owing to data limitations, the assessment underestimates the
total number of casualties associated with newly qualified drivers. However, there is also large scope for
debate about the extent to which restrictions would devalue licences and/or the affect of options in reducing
number of casualties. This reflects the limitations of the evidence base which is available on which to form
judgments.




                                                                                                               31
     Table 6.1 Summary of costs and benefits
     Area of Change                   Changes to driver training      Restrictions on learner   Restrictions on newly   Notes
                                      and testing                     drivers                   qualified drivers

     COSTS

     Pre-driver education training    Set up costs estimated £4.8m-               n/a                       n/a         Cost to the public purse
                                      £6.0m

                                      Ongoing costs (assuming take
                                      up rate of 25%) £35m per
                                      annum

     Training for ADIs to be          One off cost £4.3m-£10.7m                   n/a                       n/a         Cost to be met from the public
     compliant with new training                                                                                        purse
     regime

     Development of attitude          One off cost £1m                            n/a                       n/a         Cost to public purse.
     indicator                                                                                                          Provisional estimate of
                                                                                                                        development cost

     Learning to Drive – peer group   Ongoing costs (assuming take                n/a                       n/a         Cost to the public purse
32




     sessions                         up rate of 25%) £1.9m per
                                      annum

     Development of competency        Cost already incurred                       n/a                       n/a         Sunk cost
     framework

     Development and distribution     Ongoing costs £3.75m per                    n/a                       n/a         Cost to public purse. Assumes
     of the workbook                  annum                                                                             DSA provide a workbook to
                                                                                                                        every learner driver

     New theory test items            One off cost £1.8m                          n/a                       n/a         Cost to public purse.
                                                                                                                        Provisional estimate of
                                      Ongoing costs cannot yet be
                                                                                                                        developing new case studies.
                                      estimated

     New hazard perception test       One of cost £2m                             n/a                       n/a         Cost to public purse.
                                                                                                                        Provisional estimate of
                                                                                                                        incorporating 3D animation

     Developing new practical test    One off cost £2.1m-£5.7m                    n/a                       n/a         Cost to public purse.
                                                                                                                        Provisional estimate of
     Area of Change                    Changes to driver training   Restrictions on learner    Restrictions on newly   Notes
                                       and testing                  drivers                    qualified drivers

                                                                                                                       development cost

     Training driving examiners to     One off cost £1.9m-£5.6m                 n/a                        n/a         Cost to public purse.
     be familiar with new testing                                                                                      Provisional estimate of training
     arrangements                                                                                                      around 2000 examiners

     Potential DSA back office         One off cost £12.2m-£21.3m               n/a                        n/a         Cost to public purse.
     reorganisation because of new                                                                                     Provisional estimate of
     practical test arrangements                                                                                       aligning back office functions
                                                                                                                       to new test arrangements

     Potential increase in practical   Ongoing cost £15.1m per                                                         Cost to public purse. DSA will
     test length                       annum                                                                           require estimated additional
                                                                                                                       funds as a consequence of
                                                                                                                       driving examiners only
                                                                                                                       undertaking 6 tests per day
                                                                                                                       instead of 7. Presumption that
                                                                                                                       some of this additional cost
33




                                                                                                                       will be passed on to learners
                                                                                                                       in higher test fees.

     Minimum learning period (12                   n/a              Ongoing cost £100m-£188m               n/a         Welfare cost to learner drivers.
     months)                                                        per annum                                          The value attached to a
                                                                                                                       licence by learner drivers
                                                                                                                       unable to go on the road
                                                                                                                       legally before 12 months

     Increasing full driving age of                n/a              Ongoing cost £83m-£154m                            Hiatus of income to DSA from
     18 years old                                                   per annum                                          17 year olds but the
                                                                                                                       substantive cost is the welfare
                                                                                                                       cost to 17 year old learner
                                                                                                                       drivers from the delay before
                                                                                                                       they can possess a licence

     Minimum learning period and                   n/a              Ongoing cost £127m-£239m               n/a         Welfare cost to learner drivers
     increasing full driving age                                                                                       of operating minimum learning
                                                                                                                       period and increased full
     Area of Change                    Changes to driver training   Restrictions on learner           Restrictions on newly         Notes
                                       and testing                  drivers                           qualified drivers

                                                                                                                                    driving age

     Minimum number of lessons                      n/a             If 10 lessons - £3.3m per                      n/a              Cost to learner drivers. As yet
     with ADIs                                                      annum                                                           no agreed minimum number
                                                                                                                                    identified. Costings based on
                                                                    If 36 lessons - £94.5m per
                                                                                                                                    indicative number of learners
                                                                    annum
                                                                                                                                    requiring extra lessons

     Reduce provisional licensing                   n/a             Estimated cost of potential                    n/a              Risk of increase in accidents
     age to 16.5 years                                              additional casualty accidents -                                 involving young learner drivers
                                                                    £60.7m per annum

                                                                    Estimated cost of additional
                                                                    damage only accidents -
                                                                    £41.5m per annum

     Restrictions on Newly                          n/a                           n/a                 If 6 months: £150m-281m per   Identified costs relate to
     Qualified drivers – limitations                                                                  annum                         welfare costs to learner drivers
34




     on passengers carried and                                                                                                      associated with restrictions on
                                                                                                      If 1 year: £300m-563m per
     curfew on night driving                                                                                                        licences. There will also be
                                                                                                      annum
                                                                                                                                    opportunity costs associated
                                                                                                      If 2 years: £600m-£1,126m     with enforcing restrictions,
                                                                                                      per annum                     which are difficult to quantify.



     BENEFITS

     Non-compliance rate                           3.5%                          15%                              50%               Estimated level of non-
                                                                                                                                    compliance attached to all
                                                                                                                                    accidents for each option

     Cost savings from a reduction     Reduction in accidents and   Reduction in accidents and        Reduction in accidents and    Using DfT information a
     of accidents resulting in         casualties of 5-20%          casualties of 1-5%                casualties of 10-20%          baseline of all casualty
     casualties amongst newly                                                                                                       accidents is estimated:
                                       Total £63m-253m per annum    Total £11m-56m per annum          Total £65m-131m per annum
     qualified drivers and their
                                                                                                                                    Killed – 337
     passengers.
                                                                                                                                    Seriously Injured – 2,414
                                                                                                                                    Slightly Injured – 31,250
     Area of Change                  Changes to driver training     Restrictions on learner      Restrictions on newly        Notes
                                     and testing                    drivers                      qualified drivers

                                                                                                                              A range is applied to reflect
                                                                                                                              the uncertainty and the extent
                                                                                                                              to which interventions might
                                                                                                                              reduce casualties.
     Cost savings from a reduction   Reduction in accidents and     Reduction in accidents and   Reduction in accidents and   Applying Highways Economic
     of 'damage only' accidents      casualties of 10-40%           casualties of 1-5%           casualties of 10-20%         Note No.1 guidance to DfT
                                                                                                                              estimate of casualty accidents
                                     £85m-£339m per annum           £7m-£37m per annum           £44m-£88m per annum
                                                                                                                              a baseline of damage only
                                                                                                                              accidents is estimated:

                                                                                                                              Damage Only – 493,000

     Structured learning leads to    Ongoing savings of £51.2m                   n/a                          n/a             Cost savings to learner
     faster learning                 per annum from tests and                                                                 drivers. Assumes 25% of
                                     lessons                                                                                  learner drivers adopt the
                                                                                                                              structured learning route and
                                                                                                                              require fewer lessons to pass
35




                                                                                                                              their tests

     Structure learning leads to     Ongoing savings of £5.7m per                n/a                          n/a             Cost savings to learner
     higher pass rate                annum from test fees                                                                     drivers. Assuming 25% take
                                                                                                                              up rate amongst learner
                                     Ongoing savings of £14.9m
                                                                                                                              drivers. Due to higher pass
                                     per annum from lessons
                                                                                                                              rate learners will require fewer
                                                                                                                              tests and lessons to pass.

     Improved pass rate from non     Ongoing savings of £6.0m per                n/a                          n/a             Cost savings to learner
     structured learning             annum from test fees                                                                     drivers. Assuming marginal
                                                                                                                              improvement to pass rate will
                                     Ongoing savings of £15.6m
                                                                                                                              result in fewer tests and
                                     per annum from lessons
                                                                                                                              lessons for some learners

     Decoupling hazard perception    Ongoing savings of £5.5m per                n/a                          n/a             Cost savings to learner
     test from the theory test       annum from test fees                                                                     drivers. Learner drivers are
                                                                                                                              likely to reduce number of
                                                                                                                              attempts to pass this stage by
                                                                                                                              sitting different elements
     Area of Change                 Changes to driver training   Restrictions on learner   Restrictions on newly   Notes
                                    and testing                  drivers                   qualified drivers

                                                                                                                   individually.

     Increased feedback from test   Ongoing savings of £46.3m-               n/a                       n/a         Cost savings to learner
                                    £59.2 per annum from tests                                                     drivers. Learner drivers are
                                    and lessons                                                                    assumed to require fewer
                                                                                                                   lessons between tests and
                                                                                                                   retests from targeted
                                                                                                                   feedback.
36
Table 6.2 summarises the projected net present values of each of the options over a ten year appraisal period
using a 3.5% discount rate. 73 A breakdown of the costs and benefits of each of the options is provided in the
Annex C. For each of the options a best case scenario (minimum costs and maximum benefits) and a worst
case scenario (maximum costs and minimum benefits) are presented.

                  Table 6.2 Net Present Value - summary of options over 10 year appraisal period (£000,000s)
                  Option                        Scenario          Costs    Benefits         Net Present
                                                                                               Value

                  Option b – changes to         Best Case          509      6,451              5,942
                  driver training and
                  testing                       Worst Case         533      2,523              1,990

                  Option c – restrictions       Best Case         2,004      800              -1,204
                  on learner drivers
                                                Worst Case        3,750      160              -3,590

                  Option d – restrictions       Best Case         2,582     1,882              -700
                  on newly qualified
                  drivers                       Worst Case        4,842      941              -3,901

                  Option e – (options b         Best Case         3,092     4,949              1,857
                  and d combined)
                                                Worst Case        5,375     2,562             -2,813




Only option b shows a clear excess of benefits over costs. In all of the other cases, when net benefits are
realised will depend on how far the assumptions within the different scenarios considered prove accurate.

The benefits are heavily dependent on the accident savings which are achieved. Whilst costs are sensitive
both to the welfare value drivers attach to a licence and the presumed effect of restrictions.




73
     Per Better Regulation Executive guidance




                                                             37
                                                   Annexes

Annex A – Methodological approach used for estimating casualties
Transport Research Laboratory


1 Methodological approach
1.1 Introduction

To support the Government’s consultation on ‘Learning to drive’, estimates are required of the impact of
potential changes to driver licensing on casualties. A study was therefore commissioned to develop and
employ a methodology which could be used to provide these estimates.

The range of possible scenarios considered was:

•   For learner drivers:
       o Increasing the age at which a driver can obtain a full driving licence to 18 years;
       o Requiring a minimum learning period (12 months) for all learner drivers, implying that no driver can
           have a full licence until they are aged at least 18 years;
       o Reducing the age at which a provisional driving licence can be obtained to 16.5 years, whilst
           keeping the minimum age for a full licence at 17 years; and
       o Requiring a minimum number of hours tuition with an Approved Driving Instructor (either 10, 20 or
           40 hours).

•   For newly qualified drivers:
       o The eldest passenger should not be aged between 10 and 20 years during the first year of driving;
       o A curfew on night driving, i.e. no driving between 11pm and 5am during the first year of driving; and
       o A combination of curfew restrictions and limits on passenger carrying.

In addition, data taken from 2006 have been used to estimate the number of driver and passenger casualties
(i.e. those injured in the vehicle driven by the inexperienced driver) attributable to drivers with less than two
years experience.

Estimating the impact of changes in full licence acquisition on those casualties susceptible to these changes is
quite complex. This is because each potential change leads to a different driving population (in terms of age
and/or experience) which then has to be generated and casualty estimates derived. The basic
methodological approach employed in this study therefore relied on:
    (i)    being able to generate a table of counts for drivers by age and experience, i.e. the number of
           drivers at each age (years and months) and experience (months), and
    (ii)   a casualty prediction model.

These were used to estimate the expected numbers of killed and seriously injured (KSI) casualties resultant
with each proposed change in place. Data is based on 17-21 only, therefore represents an underestimate in
some of the scenarios; also focuses on licensed drivers only (therefore does not consider changes in numbers
of unlicensed drivers).

The approach used is discussed in more detail below, and further details of the assumptions made are given
in Appendix 1. Details of the way in which the population of drivers was estimated is given in Appendix 2 and
information on casualty estimation models used is given in Appendix 3.




                                                        A1
1.2 Calculating data on casualties involving young and inexperienced drivers

As stated above, the methodology required the generation of a table of counts of number of drivers in the
population by their age and experience. The baseline figures therefore required a table which reflected current
driver numbers (for end 2006). This was initially generated for all drivers aged 17 years 0 months to 21 years
11 months using data from a 1% sample from the Driver and Vehicle Licensing Agency (DVLA), as described
in Section 1.2.1 and in more detail within Appendix 2. This age range was selected as it includes those most
affected by the licensing scenarios being considered.

The latest available data on the number of KSI casualties (from 2006) was derived from STATS19 74 data, and
a casualty prediction model derived using information on driver age and experience, this is discussed in
Appendix 3. The prediction model was able to reproduce fairly accurately current STATS19 casualty figures
(see Appendix 3, Table 2) and is as precise as possible given the data available. As the model was applied to
both the baseline and possible scenarios, any bias inherent in the model will apply to all estimates in a
similar way.

The baseline (and other) estimates include all accidents where a young or inexperienced driver was involved
in a road accident that resulted in injuries (to themselves and/or others). The estimates include all such
accidents, irrespective of whether the young/inexperienced driver contributed to its cause.

1.2.1     Number of drivers in the population

Whilst DVLA holds data on the age of the driver and when they passed their driving test (i.e. driver data can be
calculated for how long they have held their licence and how many drivers there are at each age in the
population), it is not possible to use this data to assess the impact of proposals affecting young and
inexperienced drivers. This is because that data is only relevant at the time it is extracted from the database
as the composition of drivers is continuously changing over time, and hence any sample is a snapshot. A
methodology was therefore required to:
    (i)    estimate the numbers of drivers by their current age and experience and,
    (ii)   provide estimates of how this would change under different licensing conditions.

The first stage was to obtain counts of drivers (i.e. those with a full driving licence) at particular ages (in years
and months) and years of experience (in months). Hence, a table was generated giving counts for drivers
aged 17 years 0 months to 21 years 11 months and, for each age in months, giving a count of how many
drivers had 1 month, 2 months…. 60 months driving experience.

The number of car drivers in the current driving population was simulated using real data from a 1% sample of
drivers in the DVLA database at the end of 2006, held by TRL for research purposes. The number of drivers
by their year of age (at end of 2006) was used to simulate the number of people who become drivers at a
specific age. The simulation process was adjusted as necessary to obtain a good agreement with known data.
Table A.1 shows the agreement between the simulated figures from the model and the 2006 DVLA data
obtained from the 1% sample. The number of drivers obtaining full driving licences varies over time and so
modelling the population of current drivers has to try to allow for this. This means that it is not a simple model.




74
  STATS19 is the database of information on reported injury accidents occurring on British roads, containing data on accidents, the vehicles involved
and injured casualties.




                                                                          A2
                     Table A.1 Count of driver numbers from DVLA 1% sample and simulated table

                                             No. of licensed drivers
                                              of this age at end of                Counts from simulated
                       Age of driver              2006 – DVLA                         table – model                       % difference
                       17 years                       78,700                               78,743                            -0.1%
                       18 years                       292,700                             292,693                             0.0%
                       19 years                       385,000                             386,472                            -0.4%
                       20 years                       422,100                             421,954                             0.0%
                       21 years                       452,700                             451,916                             0.2%

This table of driver counts was subsequently expanded beyond the 17-21 year age range up to the age of 60
years old, for more general use.

Changes to the licensing system will result in changes to the number of drivers in each cell of the simulated
population table, e.g. if full licence acquisition is not possible until drivers are aged 18 years or more, then
there could not be any 17-year old drivers in the population table. The simulated table of driver counts by age
and experience was therefore adjusted to reflect each licensing scenario being investigated. The impact on the
population of drivers varied, depending on the scenario under consideration.

In general, the figures used to generate the baseline population table of drivers provided the initial numbers.
The numbers of drivers (by age) for each scenario were shifted (in time), such that they were older and/or had
less experience and such that the driver population then corresponded with the licensing proposal being
considered. So that, for example, if drivers have to be at least 18 years of age before they can acquire a full
licence then the 17-year old drivers would be added to the current 18-year olds, but they would be older than
under the current system. It is assumed that all the current 17-year olds would still acquire licenses and so be
in total population of drivers, but they would be older and with less driving experience – factors which will
influence their subsequent accident liability. There are clearly many ways in which these 17-year olds can be
added into the population of drivers.

1.2.2       Casualty estimation

The number of casualties associated with drivers (aged from 17 to 21 years), in the simulated table was
determined by applying a casualty liability model, details of which are given in Appendix 3. This model could
only take into account the age of the driver and how much driving experience they had gained since acquiring
a full driving licence, since this is the only information available (for example, information on exposure was not
available) 75 . The casualty liability per driver (the number of casualties associated with each driver within a
year) was calculated from the model and then scaled by the numbers of drivers in each age /experience cell to
provide the estimated numbers of casualties for that cell. The parameters in the model were adjusted such that
the estimated casualty numbers under the current licensing scenario closely reflected the observed number
from the STATS19 database.

It is important that this simulation process reflects the actual numbers of casualties as recorded within
STATS19 under the current licensing regime; only by being able to estimate current numbers of casualties can
any changes to the licensing regime be estimated with confidence.




75
     Accident liability depends on exposure, i.e. number of miles driven, but this is not generally available at driver level.




                                                                                A3
To develop the statistical model to estimate the accident liability of drivers based on their age and experience,
data from a Cohort II 76 study (Wells et al., 2008) was used. The model was adjusted to reflect the observed
STATS19 casualty numbers for 2006. The resultant model intrinsically incorporated an accident to casualty
adjustment factor.

As explained in Section 1.2.1 of this Annex, each of the licensing scenarios of interest would be expected to
alter the table of driver numbers (Table A.1) in some way. This variation in the number of drivers resulted in a
change in the numbers of predicted casualties, mainly because the base number of drivers had changed. The
overall age and experience of the driver population also changed to some degree, which contributed to a
further change in the estimated numbers of casualties. The impact of a licensing scenario on numbers of
casualties was judged by comparing the resulting modified number of casualties with the baseline data. As
already mentioned, whilst estimates derived from using the casualty prediction models may be biased, the bias
will be similar in each scenario. Hence the difference between estimates will eliminate this potential
inaccuracy.

Double counting (which could result if more than one young driver was involved in a collision, thereby leading
to casualty counts being ascribed to both drivers) was avoided by analysing STATS19 (2006 data) for single
vehicle, two-vehicle and multiple-vehicle tables of casualties by age of driver(s) separately. Single vehicle
collisions cannot involve double counting. Tables for two-vehicle collisions were generated giving the ages of
both drivers and hence it was possible to identify casualties associated with 17 year old drivers (or 18, 19 etc.)
and so avoid double counting. An adjustment was also made to allow for multiple counting of casualties where
there were three or more vehicles involved in a collision.

It is important to state that when considering new driver restrictions full compliance is unlikely. Studies on
illegal drivers (Knox et al., 2003) and on new driver compliance to driving restrictions (Goodwin et al, 2006)
provide some information which could be used to adjust the casualty estimates. However, this data is very
limited and most of the research to date has been carried out in other countries. Therefore, given the lack of
available data on this subject, it has been assumed in the estimates provided that there is total
compliance (i.e. 100%). Clearly this will not be the case and the model does allow for partial compliance
once more robust data becomes available (see Section 5). The partial impact assessment applies some
assumptions about compliance rates in an attempt to manage this (see Section 4.0 of the partial impact
assessment).

A second approach was used where an adjustment allowing for the fact that not all accidents involving young
or inexperienced drivers are the result of their actions. STATS19 contains information on those factors likely
to have contributed to the cause of the accident. This data on contributory factors was used to identify those
accidents where the behaviour of the young/inexperienced driver was judged likely to have contributed to the
cause of the accident, for example ‘failed to look properly’, or ‘loss of control’ etc. The casualty estimates were
then adjusted to take into account the proportion of accidents where the driver was not judged to be a
contributory factor. This second approach assumed that the accident still happens with different licensing
scenarios but not involving the young/inexperienced driver (see Section 5 in this Annex).

Given that the count of drivers for each combination of age and experience changes according to the licensing
scenario of interest, the impact of this change provided a revised casualty estimate which, when compared to
the current licensing scenario, gives an indication of impact on susceptible casualties. As mentioned
earlier, the main focus of this simulation was on young drivers, which was appropriate for most of the licensing
scenarios being considered. However, it does means that, where a licensing scenario also affected older new

76
   The Cohort II study was a large-scale questionnaire survey of learner and new drivers. A new “cohort” of drivers who had recently taken a practical
driving test was approached every three months, thereby building a sample taken from 16 time-points over four years. Those drivers who passed the
practical test were then followed-up at six months, 12-months, 24-months and 36-months after their test. They were asked to complete a follow-up
questionnaire at these times which included questions about accidents.




                                                                          A4
drivers, the estimated casualty saving will be an under-estimate because the analysis only considers drivers
aged 17 to 21 years 77 .

1.2.3       Assumptions associated with all licensing scenarios

The impact on casualty numbers of each of the scenarios of interest is outlined in Section 3 in this Annex,
together with any associated assumptions. There are some overall assumptions which apply to all the
scenarios:
   o The table of driver numbers is sufficiently representative of actual drivers in the population and so
       provides a mechanism for simulating the numbers of casualties;
   o Using a table for drivers aged between 17 and 21 years is adequate for estimating the impact of the
       scenarios specified relating to changes in the licensing regime;
   o A separate accident liability model derived for killed and seriously injured casualties reflects the
       observed STATS19 reported casualties with sufficient accuracy;
   o The accident liability model is applicable to all licensing scenarios under consideration;
   o The methodology assumes that the new licensing scenario has been operational for some time, and so
       any transition period is not of relevance;
   o The approach focuses on the impact on those currently in the system (i.e. licensed drivers, not those
       who choose to drive unlicensed);
   o The approach assumes total compliance (see Section 5 in this Annex, on Limitations);
   o One approach includes all accidents involving a young or inexperienced driver and a second approach
       takes into account whether their actions contributed to the cause or not, (see Section 5 on Limitations).

Further assumptions are outlined in Appendix 1 and details of the casualty model are given in Appendix 3.


2 Estimating killed and seriously injured casualties associated with
  inexperienced drivers
An estimate of the number of KSI casualties associated with inexperienced drivers was required.
Inexperienced drivers are defined here as those with up to two years driving experience. Ideally, this
analysis would have been undertaken for all casualties associated with inexperience, but this is problematic
and complex to calculate if double, triple or even quadruple counting (i.e. more than one inexperienced driver
is involved in the collision) in multiple vehicle accidents is allowed for. Hence, to avoid double counting, an
estimate was generated relating only to the inexperienced driver and/or their passengers.

Two pieces of information were used to estimate the number of inexperienced drivers plus their passengers
who are killed or seriously injured in road accidents. First, counts of drivers and their passengers killed or
seriously injured by the age of the driver were derived from STATS19 data. Secondly, the proportion of car
drivers with up to two years of driving experience (by the age of driver) was derived from the table of counts of
drivers in the population (under the current licensing regime).

A table of counts of drivers, with rows and columns for driver age and experience, was constructed as
described in Appendix 2; it was extended to include drivers up to 60 years of age. It was assumed that the
number of drivers who were aged over 60 years and had less than two years driving experience was
negligible. The proportion of drivers with less than two years of driving experience for each age was computed
by summing just those drivers with up to two years of experience. The proportion of drivers with less than two
years of driving experience was multiplied for each driver age by the observed number of ‘driver + their
passenger’ numbers as derived from the STATS19 database. The result represents the likely number of

77
     Potentially this means that the figures associated with scenarios 2 and 4 are likely to be under-estimated.




                                                                              A5
‘driver + their passenger’ fatalities or KSI casualties for those accidents involving an inexperienced driver (see
the calculation in Table A.2).

            Table A.2 Proportion of drivers with up to two years experience and calculation of
            associated casualties

                                                                              Estimated casualties
                                                                          (inexperienced drivers and
                            All drivers and their        Percentage         their passengers) where
                           passengers by injury            with <2yr        driver had less than two
                            severity (STATS19)            experience            years experience
                                                       (from model of
               Driver                                        driver
                age          Fatal          KSI          population)          Fatal           KSI
                 17            83           564             100.0%              83            564
                 18           100           861             100.0%             100            861
                 19            88           764             79.6%               70            608
                 20            72           659             30.6%               22            202
                 21            84           547             15.0%               13             82
                 22            66           462             13.8%               9              64
                 23            53           393             13.6%               7              54
                 24            47           387             13.0%               6              50
                 25            48           400             11.8%               6              47
                 …             …             …                 …                …              …

The following table, Table A.3, shows the estimated number of casualties associated with drivers and their
passengers for inexperienced drivers and for drivers of all ages, (i.e. from 17 to 60 years).

            Table A.3 Estimated numbers of casualties (inexperienced driver plus their passengers)
            associated with drivers who have no more than two years of driving experience

                                                    Count of casualties
                                                  where driver <2yr driving      Count of casualties for all
      Severity                                          experience                 drivers aged <61yrs
      KSI – ‘driver + their passengers’                     2751                           11782
      Fatal – ‘driver + their passengers’                    337                           1282

This suggests that about 23% of all KSI ‘driver plus their passenger’ casualties are associated with
inexperienced drivers, and that about 26% of fatal casualties are similarly associated.


3 Estimating casualties relating to changes in licence acquisition
This section presents changes in casualty estimates, relative to the baseline, i.e. they represent the reduction
to be expected in those casualties susceptible to the proposed change, in light of the various limiting
assumptions. As mentioned above, these estimates cannot, by their nature, be precise and therefore the 95%
confidence interval has also been provided for each estimate, meaning that we expect the estimated value to
lie within the stated range 95 times out of a hundred.

The casualty estimates provided in this section assume that the revised population of drivers will be fully
compliant, e.g. 17 year olds will take their driving test at 18 years rather than choose to drive unlicensed.




                                                            A6
Clearly this will not be the case for all those affected by the changes, but as already mentioned there is
insufficient data on the likely levels of compliance – see Section 5. It is important to note that casualty
savings will be lower if drivers are not fully compliant.

Equally, there are many accident situations where the young driver did not contribute to the cause. The
casualty estimates have attempted to take this into account.

3.1 Scenario 1: Increasing the age at which a full driving licence can be obtained to 18 years

The first scenario considered was that no driver can acquire their full car licence until they are aged 18 years.
It follows that there will no longer be any drivers aged 17 years in the driver population. There are various
ways of modelling this effect, depending on the assumptions made about the age at which current 17-year
olds pass their test when they reach 18. These assumptions lead to a slightly different driving population and
hence lead to different casualty estimates. A range of assumptions was applied to the data to determine the
range of effect likely. The two situations below represent the current situation (albeit the drivers are one year
older) and a likely alternative.

Situation (a): Increasing the full driving age to 18 years, assuming current 17-year olds have similar learning-
to-drive experiences when they pass their test at 18 years of age

It was assumed that the distribution of 17-year olds who now have to wait until they are 18 years of age is the
same as it would have been under the current licensing regime, see Figure A1. There will be no drivers aged
17 years in the driver population. As explained earlier, the expected number of KSI casualties was estimated
using the resulting modified Table A.1 together with the accident liability model.

Finding: Given the assumptions listed in Section 1.2.3 of this Annex, it was an estimated that this measure
could affect 430-700 KSI casualties (95% confidence interval), which represents between 7.3% to 11.8% of
the total population potentially susceptible (i.e. the casualties associated with drivers aged 17 years). This is
likely to be an overestimate because it assumes full-compliance.

Situation (b): Increasing the full driving age to 18 years, assuming current 17-year olds take their test early in
their 18th year, and that they will have been preparing for it while still aged 17 years.

As an alternative to the assumption in situation (a), here it was assumed that the distribution of 17-year olds
who now have to wait until they are 18 years of age is ‘front-loaded’ relative to what it would have been under
the current licensing regime, i.e. drivers are passing their test earlier in their 18th year of age than they would
have in their 17th year of age. Hence, the driver population adopts a specific distributional form for the extra 17-
year olds added to the current count of 18-year olds.

The distribution of those 17-year olds has arbitrarily been assumed to have a ‘front-loaded’ form, as modelled
by using a Gamma distribution, see Figure A2 (scaled to reproduce the sample size of 17-year old drivers) –
there is no evidence to support this other than, if the option of passing the test earlier is to be considered, it
appears an appropriately shaped distribution. The distribution of 17-year olds passing their test is shown by
Figure A1 and the Gamma function assumed when ‘front-loading’ the age at passing is given in Figure A2.
Adding these 17-year olds to the current 18-year olds provided a revised table of driver numbers which only
affected new 18-year old drivers.

Finding: Given the assumptions listed in Section 1.2.3 of this Annex, it was estimated that the measure could
affect 310-590 KSI casualties (95% confidence interval); this represents between 5.2%-9.8% of the total
susceptible population. These values are lower than for situation (a) because the now 18-year old drivers




                                                        A7
become drivers sooner and so are slightly younger. This is likely to be an overestimate because it assumes
full-compliance.


                    2500                                                                                 3000




                                                                                 front-loaded 17-years
                    2000                                                                                 2500
  17-year drivers




                                                                                                         2000
                    1500
                                                                                                         1500
                    1000
                                                                                                         1000
                    500
                                                                                                         500
                      0                                                                                    0
                           1   2   3   4   5   6   7   8   9   10   11   12                                      1   2    3   4   5   6   7   8   9   10   11   12
                                               month                                                                                  month


Figure A1. Current distribution of age of 17-                                                                            Figure A2. Hypothetical Gamma
year olds when passing test (by month of                                                                                 distribution for age of 17-year
age)                                                                                                                     olds when passing their test,
                                                                                                                         assuming front loading
Source: Cohort II data

3.2 Scenario 2: Requiring a minimum learning period of 12 months

It was assumed that the date at which a person acquires their provisional licence signifies the start of their
learning period. If they have to hold this provisional licence for at least 12 months to ensure they have at least
12 months of learning, then this will, in effect, lead to there being no 17-year old drivers in the population. The
distribution of time taken from obtaining a provisional licence to taking the practical driving test was derived
from the previously mentioned Cohort II study (Wells et al., 2008). This provided an indication of what
percentage of new drivers takes less than 12 months of tuition. Table A.4 shows the percentage of drivers who
take less than a 12-month learning period by age of driver. (Under the proposed scenario, it was assumed that
all new drivers who were still aged 17 years would have to take 12 months of learning).

                               Table A.4 Percentage of drivers who said that they took less than 12-months from
                               provisional licence to passing their driving test (Wells et al., 2008)

                                                                                   Age at time of taking test
                                                                               18       19          20        21
                                                       % <=12mths             20.7%      6.9%       7.3%       8.2%

In order to generate the modified table of driver counts, a percentage of the distribution who took less than 12-
months to learn was shifted to having passed their test later in time. In practice, this meant that all 17-year olds
were shifted, 21% of 18-year olds were shifted and 7% of 19, 20 and 21-year olds were shifted. The number of
drivers in each age band ‘shifted’ to a later time period was proportional to the current number of drivers within
the age group, i.e. if there were 1,000 18-year olds then 207 (20.7%) were ‘shifted’ to passé their test when 19
years of age.

This approach may overestimate the effect on the counts of drivers by age and experience in the modified
table, since it is quite likely that they will obtain their provisional licence earlier than currently even if they do
not start to learn earlier.

Finding: Given the assumptions listed in Section 1.2.3 of this Annex and that learner drivers are not likely to
apply for their provisional licence earlier, it was estimated that this measure could affect 580 to 840 KSI
casualties (95% confidence interval); this represents 9.7% to 14.1% of the total susceptible population. These




                                                                                                                A8
figures are only adjusted for casualties associated with drivers aged from 17 to 21 years. The effect will be
greater if applied to the whole population of new drivers.

3.3 Scenario 3: Reducing the age at which a provisional licence can be obtained to 16.5 years

It was assumed that people can still obtain a full driving licence aged 17 years, but that they can obtain a
provisional licence when 16.5 years. This then ensures that any 17-year olds coming to test when they are
just 17 have the chance to have a 6-months learning period. It was assumed that this would then allow some
17-year olds to pass their test earlier than they would under the current system.

The simulated driving population was adjusted by ‘front-loading’ the numbers of 17-year olds relative to the
current numbers (i.e. it was assumed that a greater proportion of 17-year olds pass their test earlier in their
17th year of age). There are no data to justify any specific adjustment being used and so the arbitrary Gamma
function, Figure A2, used within Scenario 1, situation (b) was applied.

Finding: Given the assumptions listed in Section 1.2.3 of this Annex and that 17-year olds would try to pass
their test soon after they turn 17, it was estimated that this scenario would result in more younger drivers,
which in turn may increase the KSI casualties (95% confidence interval from an increase of 280 to a reduction
of 24); this corresponds to a 95% confidence interval from minus 4.7% to plus 0.4%. The increase in
casualties occurs because some 17-year olds are assumed to become full licence drivers earlier in their
driving career and hence are younger when they start driving.

3.4 Scenario 4: Requiring a minimum number of hours of tuition with an approved driving instructor

The average number of hours of professional tuition received by candidates when preparing for their test was
derived from Wells et al., (2008). Any relationship between accidents and hours of tuition may be attributable
to a number of factors. An analysis of road accidents in the Cohort II study, controlling for age, experience, sex
and measures of exposure found when hours of professional tuition was added to the model no significant
extra explanatory value (Wells et al., 2008). Hence, the effect of a minimum tuition period may simply be to
delay the time at which a candidate can take their test.

Wells et al. (2008) found that 2.2% of candidates choose to take less than 10 hours, 9.7% choose to take less
than 20 hours and 52.1% choose to take less than 40 hours of professional tuition. Data from this study
suggest a relationship between the hours of professional tuition and the number of months from provisional
licence to taking the test; about 2.4 hours professional tuition accrue for every extra month before taking the
test. This is an approximate relationship which does not take into account other factors, such as the number of
hours of practice taken with friends or relations. It is therefore suggested that estimates derived using this
relationship should be viewed with caution.

It was calculated that, in order to achieve the necessary 10, 20 or 40 hours of professional tuition respectively,
an additional 0.05, 0.31 and 3.14 months of learning, on average, would be required. In order to evaluate
these three options, the distribution of drivers in the simulated population needed to be shifted in time by these
amounts and revised estimates of KSI casualties obtained. Shifting the learning time by 0.05 months made
very little difference, but shifting it by just over 3 months seemed to affect more casualties relative to a ‘no-
change’ scenario.

Finding: Given the assumptions listed in Section 1.2.3 of this Annex and the approximate relationship
between hours of tuition and time to practical test, it was estimated that there would be a reduction in KSI
casualties, corresponding respectively to a requirement of 10 hours, 20 hours or 40 hours professional tuition
being taken. The 95% confidence intervals for the number and percentage of casualties are, respectively:




                                                        A9
          •    10 hours        -140 to 155 (-2.4% to 2.6%),
          •    20 hours        -110 to 190 (-1.8% to 3.2%)
          •    40 hours         260 to 530 (4.3% to 9.0%).

3.5 Summary of casualty estimates relating to licence acquisition

Table A.5 gives estimated casualty reductions arising from the proposed measures, where casualties are
those susceptible to the proposed measures. It should be noted that the figures provided assume full
compliance, and therefore the true figures will be less than those provided. They also include all accidents
involving young and inexperienced drivers, not just those where their actions may have contributed to the
cause of the accident.

Table A.5 Estimated reduction in the number of casualties as a result of proposed measures for full
licence acquisition

Figures assume full compliance and include ALL                                           95% confidence interval
accidents involving young/inexperienced drivers (i.e.
drivers aged between 17-21years)
                                                                                    Casualties               Percentage
                                                                                    affected by              affected by
Proposed measure                                                                   each scenario            each scenario
•   Increasing the age at which a driver can obtain a full
                                                                                       430 to700            7.3% to 11.8%
   driving licence to 18 years
• Increasing the age at which a driver can obtain a full
                                                                                      310 to 590             5.2% to 9.8%
   driving licence to 18 years, passing test earlier
• Requiring a minimum learning period (12 months) for all
                                                                                      580 to 840            9.7% to 14.1%
   learner drivers
• Reducing the age at which a provisional driving licence
   can be obtained to 16.5 years, whilst keeping the                                  -280 to 24             -4.7% to 0.4%
   minimum age for a full licence at 17 years
• Requiring a minimum number of hours tuition with an
                                                                                      -140 to 155            -2.4% to 2.6%
   approved driving instructor (10 hours)
• Requiring a minimum number of hours tuition with an
                                                                                      -110 to 190            -1.8% to 3.2%
   approved driving instructor (20 hours)
• Requiring a minimum number of hours tuition with an
                                                                                      260 to 530             4.3% to 9.0%
   approved driving instructor (40 hours)
Note: Negative numbers imply that there would be an increase in number of estimated casualties, if the confidence interval spans zero it implies that the
casualty reduction is not statistically significant.

Using STATS19 contributory factor data, the numbers of casualties affected by each scenario are reduced by
removing those accidents where the young/inexperienced driver was judged not to have contributed to the
cause of the accident, i.e. it is assumed that the accident would still have happened. This then gives the
estimates shown in Table A.6, relating to those casualties susceptible to the proposed measure, adjusted for
contributory factors.

Table A.6 Estimated reduction in the number of casualties as a result of proposed measures for full
licence acquisition, adjusted for contributory factors

Figures assume full compliance to the proposed measure,                                  95% confidence interval
but ONLY include those accidents where the
young/inexperienced driver may have contributed to the
cause of the accident
                                                                                     Casualties              Percentage
Proposed measure                                                                     affected by             affected by




                                                                          A10
                                                                                   each scenario            each scenario
•    Increasing the age at which a driver can obtain a full
                                                                                      320 to 590            5.3% to 10.0%
     driving licence to 18 years
•    Increasing the age at which a driver can obtain a full
                                                                                      220 to 500             3.8% to 8.5%
     driving licence to 18 years, passing test earlier
•    Requiring a minimum learning period (12 months) for all
                                                                                      430 to 700            7.2% to 11.7%
     learner drivers
•    Reducing the age at which a provisional driving licence
     can be obtained to 16.5 years, whilst keeping the                                -250 to 50            -4.2% to 0.8%
     minimum age for a full licence at 17 years
•    Requiring a minimum number of hours tuition with an
                                                                                      -140 to 150           -2.4% to 2.6%
     approved driving instructor (10 hours)
•    Requiring a minimum number of hours tuition with an
                                                                                      -120 to 180            -2.0% to 3.0%
     approved driving instructor (20 hours)
•    Requiring a minimum number of hours tuition with an
                                                                                      170 to 450             2.8% to 7.6%
     approved driving instructor (40 hours)
Note: Negative numbers imply that there would be an increase in number of estimated casualties, if the confidence interval spans zero it implies that the
casualty reduction is not statistically significant.




4 Estimating casualties relating to post-licence restrictions
Two restrictions have been considered, together with a combination of the two post-licence restrictions. The
estimates provided below give the number and percentage change in casualties in relation to the baseline, i.e.
they represent the estimated reduction in those casualties susceptible to the proposed measure. As with
Section 3in this Annex, it was assumed that all drivers would comply with the proposed restrictions. Clearly
this will not be the case, however data on the extent to which drivers can be expected to comply is currently
insufficient (see Section 5 in this Annex). The figures provided therefore represent an overestimate of the
likely effect.

It is quite likely that drivers may adopt displacement strategies, i.e. within a time curfew they may make the
journey at a different time, or two new drivers may each drive, rather than share a car if there are age
limitations on passengers. These are discussed in Section 5. The casualty estimates provided do not allow for
the impact of these strategies as there is insufficient information on what proportion of drivers will respond in
which ways to the proposed measures.

4.1 Scenario 5: The age of the eldest passenger should not be between 10-20 years in the first year of
    driving

Data from Wells et al. (2008) shown in Table A.7, shows that, for drivers in their first year of driving, 27% of all
KSI accidents occur when the eldest passenger is in the 10-20 year age category. The estimated number of
KSI casualties was calculated assuming that 27% (from Table A.7) of existing casualties would not occur if
their eldest passenger was not in the 10 to 20-year old age range. It is recognised that this represents the
maximum reduction in the number of casulaties susceptible to the proposed restriction because it is probable
that compliant drivers would take an older passenger, drive on their own or not do the journey. Taking an
older, or no passenger, may reduce the risk of the accident happening and it was assumed, for the derived
estimated reduction, that the accident would not occur.

                Table A.7 Number of accidents where the eldest passenger was aged 10-20 years,
                tabulated by accident severity (Wells et al., 2008)

                                                 Accidents during the first year of driving




                                                                          A11
                                                 Age of oldest passenger

                                                <10 or >20yrs     10-20yrs        Total

                               No injury             2267           754           3021
                              Slight injury           206            82            288
                                   KSI                72             26            98
                                 % KSI               73%            27%             -
                                  Total              2545           862           3407


Finding: Given the assumptions listed in Section 1.2.3 of this Annex and it is assumed that drivers are fully
compliant in not having an eldest passenger aged 10-20 years, then it is estimated that this measure will affect
650 to 910 KSI casualties (95% confidence); this represents 10.9% to 15.2%. This is likely to be an
overestimate.

4.2 Scenario 6: No driving between 11pm and 5am in the first year after full licensure

Table A.8 gives the 2006 STATS19 data showing the percentages of KSI casualties that occurred during the
proposed curfew and non-curfew periods. The data are recorded in whole hours and it was assumed that the
curfew period would result in less casualties occurring between 11pm and 5am .The percentage of KSI
casualties varies little by the age of the driver. This curfew scenario will only apply in the first year of driving
and so it was first necessary to estimate the modelled number of casualties which occur during the first year
and to reduce these by the percentages given in Table A.8.

It was assumed that there was no displacement of the trip/journey as there is little data on this (see Section 5
in this Annex). Perhaps more realistically, it would be better to assume that the trip still occurs but is brought
forward in time, so as to beat the curfew. However, the accident liability model does not include the time of day
as a factor, which may have been useful when adjusting for the effect of the curfew, as accident liability tends
to be higher during the proposed curfew period. This approach is likely to overestimate the number of
casualties affected.

            Table A.8 Proportion of KSI casualties occurring during the proposed curfew period by age
            of driver, (source: STATS19 data)

                                  Driver’s    % of KSI casualties occurring from
                                    age        6am to 10pm         11pm to 5am
                                     17             68%                32%
                                     18             62%                38%
                                     19             64%                36%
                                     20             62%                38%
                                     21             64%                36%

Finding: Given the assumptions listed in Section 1.2.3 of this Annex and if it is assumed that all drivers are
fully compliant (although very unlikely), then it is estimated that this measure will affect 930 to 1180 KSI
casualties (95% confidence interval); this represents 15.7% to 19.8% of the population susceptible to this
proposal. The figures are likely to be an overestimate.

4.3 Scenario 7: A combination of curfew restrictions and limits on passenger carrying




                                                         A12
The data for Scenarios 5 and 6 give can be combined to provide an estimate of the number of casualties
affected if both were applied together. The estimates from both scenarios were combined, but adjusted to take
into account the overlap, (i.e. casualties arising from accidents between 11pm and 5am when the eldest
passenger was between 10-20 years old). These estimates will very much be an upper limit, both because
they assume full compliance, which is unlikely and because drivers may on occasion violate the restriction (i.e.
they will not always be in violation).

Finding: Given the assumptions listed in Section 1.2.3 of this Annex it is estimated that this measure will
affect 1450 to 1670 KSI casualties (95% confidence interval); this represents 24.4% to 28.1% of the
susceptible population affected by this proposal. The figures are likely to be overestimated.

4.4 Summary of estimates relating to driver restrictions

Table A.9 gives estimated casualty reductions arising from the proposed restrictions. It should be noted that
the figures provided assume full compliance, and therefore represent overestimates of the true effect.
Equally, they include all accidents involving inexperienced drivers, irrespective of whether the driver
contributed to the cause of the accident.

Table A.9 Estimated reduction in the number of casualties as a result of restrictions on newly qualified
drivers

Figures assume full compliance                                          95% confidence interval
                                                                    Casualties        Percentage
                                                                    affected by    affected by each
Proposed measure                                                   each scenario       scenario
• The eldest passenger not aged between 10 and 20 years
                                                                    650 to 910      10.9% to 15.2%
   during the first year of driving
• A curfew on night driving, i.e. no driving between 11pm and
                                                                    930 to 1180     15.7% to 19.8%
   5am during the first year of driving;
• A combination of curfew restrictions and limits on passenger
                                                                   1450 to 1670     24.4% to 28.1%
   carrying

Using STATS19 contributory factors data, those accidents where the inexperienced driver was judged unlikely
to have contributed to the cause of the accident were excluded. Estimates relating to those casualties
susceptible to proposed measures, adjusted for contributory factors are shown in Table A.10.

Table A.10 Estimated reduction in the number of casualties as a result of driver restrictions, adjusted
for contributory factors

Figures assume full compliance to the proposed measure,                95% confidence interval
taking into account where the young driver did not have
contributory factor, i.e. perhaps more realistic
                                                                      Casualty       Percentage
Proposed measure                                                     reduction        reduction
• The eldest passenger not aged between 10 and 20 years
   during the first year of driving                                  440 to 710     7.4% to 11.9%

•   A curfew on night driving, i.e. no driving between 11pm and
                                                                    650 to 910      10.9% to 15.3%
    5am during the first year of driving;
•   A combination of curfew restrictions and limits on passenger
                                                                   1050 to 1290     17.6 to 21.6%
    carrying




                                                          A13
5 Limitations
The casualty prediction model used to determine the casualty estimates is based on a number of assumptions
and these have been detailed earlier. There are three key limitations to the casualty estimates provided.
These are discussed below and arise because the approach taken relies on being able to model the
population of drivers once the proposed measures are introduced, and this requires certain assumptions being
made.

 The limitations relate to:
• assumed levels of compliance,
• displacement,
• accidents in which the young/inexperienced driver is likely to have not contributed to the cause of the
   accident.

No attempt has been made to allow for compliance or displacement activities in the estimates given as
the data is too limited (the model might be able to take these into account once more robust data becomes
available). Some of the figures provided do, however, take account of the contribution of drivers to the cause
of the accident.

5.1 Compliance

It is clearly the case that a proportion of people will choose to drive unlicensed if the age at which they can
acquire a full driving licence is increased. Similarly, some drivers will choose to ignore any restrictions placed
on them. The casualty estimates provided above do not take account of compliance rates as there is
insufficiently detailed information available to provide reasonably accurate figures.

It has been established that unlicensed drivers have a higher accident liability than do licensed drivers (Ward
et al., 2008; Clarke et al., 2007), and it could reasonably be assumed that they are more likely to violate other
traffic regulations such as speed limits. Data on the extent to which people might violate the proposed
measures is limited to findings by Knox et al. (2003), Clarke et al. (2007) who found that seven per cent of
their crash involved sample involved drivers without a full driving licence; and Ward et al. (2007) who suggest
that full driver licence holding is in decline and that unlicensed young drivers tend to be low-income workers,
socially excluded individuals and joy-riders, and have between 2.7 and 9 times higher crash risk than for all
drivers.

Knox et al. (2003), in DfT Road Safety Report No 48 ‘Research into Unlicensed Driving’, reported on a series
of interviews with members of the general public. They found that 21% of drivers had driven without a licence
and 22% of drivers had at some time driven outside the regulations when a provisional licence holder. The
study also found that, for drivers aged under 18 years of age, 20% had driven without a licence. This is the
percentage that had, at some time, driven without a licence – which is not the percentage on the road at any
one time.

This percentage could be applied to the casualty prediction model under the (strong) assumption that the
estimates provided above could not be realised because some (up to 20%) of drivers would be non-compliant.
Non-compliant drivers are quite likely to be those who would speed or drive in a more risky way, and hence
would contribute to a higher proportion of accidents as compared to those who are compliant, but would not be
responsible for all accidents. At best, any adjusted estimates of casualty savings using the 20% figure would
provide a lower limit of potential reductions in casualties, i.e. it would reduce the estimated number of
casualties related to the susceptible population of road users by the maximum amount.




                                                        A14
There is evidence from other countries that new drivers within a graduated licensing regime do not fully comply
with the conditions imposed on them. Goodwin et al. (2006), reporting on their findings in North Carolina,
suggest in the introduction to their paper that 80% of probationary drivers may at times violate the graduated
licence scheme restrictions. They quote studies conducted in New Zealand, Nova Scotia, and California where
it was found that as many as 50% of young drivers report violating night-time restrictions and up to 80%
acknowledge at least occasionally violating passenger restrictions (Begg et al., 1995; Mayhew et al.,1998;
Williams et al., 2002). Goodwin et al. (2006) also found that 22% of teenage drivers violated night-time
restrictions with or without parental permission and 45% violated passenger restrictions with or without
parental permission. Taken together, 52% violated either passenger and/or night-time restrictions with or
without parental permission.

It is accepted that these numbers violating are within a graduated licensing system and that these new drivers
do not violate all the time. However, these figures provide an indication of how curfew and passenger
restrictions may not be fully complied with and so will not provide the maximum size of effect. Given that the
British licensing system is not graduated and there are other differences between Britain and the countries
where this research was conducted, it is difficult to generalise from these reported findings to the scenarios
being currently considered.

Given the lack of robust data on compliance, the estimates provided assume full compliance. However, the
method adopted here allows a figure on levels of compliance to be taken into account, once such data
becomes available.

5.2 Potential for displacement activities

A new (and restricted driver) may be thwarted in not being able to make a planned journey. This may, in turn,
lead to a displacement activity. There are a number of potential displacement activities which the new driver
may adopt, such as:

•   For night-time restrictions
       o Reschedule the journey to a non-curfew period
       o Do not make the journey
       o Make the journey but not as a driver.

•   For passenger restrictions
       o Do not take any passenger in the 10 year to 20 year age group
       o Take an older passenger
       o Two new drivers take two cars rather than share.

Potentially such displacement activities will impact on the casualty estimates provided. If they do not drive,
then they cannot be accident involved in the same way. However, it is very likely that a number of new drivers
would not comply as was discussed above.

There are no reliable data on the likelihood of new drivers adopting these alternative strategies and so the
estimates provided make no allowance for displacement activities.

5.3 Accidents where the young/inexperienced driver is likely to have contributed to the cause of the
    accident

Young and/or inexperienced drivers will not contribute to the cause of all accidents they are involved in. Data
are recorded at accident scenes which assign contributory factors to all those drivers involved. An analysis of
contributory factors, as recorded in the 2006 STATS19 database, looked at accidents by driver’s age (in the




                                                       A15
range 17 years to 21 years). If road users (other than the young/inexperienced driver) were judged as very
likely to have contributed to the cause of an accident then these accidents were deemed as not having been
affected by the age or experience of the driver.

Using this data, a table was constructed giving the percentage of accidents where the young/inexperienced
driver was judged not to have contributed to its cause. This is given as Table A.11, below. The percentages in
this table were used to adjust the numbers of casualties associated with the susceptible road users under the
assumption that these accidents, not attributable to the young/inexperienced driver, would still have happened.
It was further assumed that the number of casualties affected were directly proportional to the numbers of
accidents affected. This approach was used in the summary tables provided for each of the proposed
measures.

           Table A.11 Proportion of accidents where the actions of the young/inexperienced driver
           were judged not to have contributed to the cause of the accident

                                                                                  Percentage
                                                               All accidents
                                                                                     where
                                                                 involving
                                              Contributory                      young/inexperi
                 Age of     Contributory                        young driver
                                            factor assigned                       enced driver
                 young    factor assigned                       with at least
                                             to other road                       judged not to
                 driver   to young driver                           one
                                                  user                               have
                                                                contributory
                                                                                contributed to
                                                              factor recorded
                                                                                   the cause
                   17           424               120              544                22%
                   18           681               230              911                25%
                   19           603               262              865                30%
                   20           545               264              809                33%
                   21           483               250              733                34%




                                                        A16
References
Begg D J, Langley J D, Reeder A I & Chalmers D J (1995). The New Zealand graduated driver licensing system:
Teenagers' attitudes towards and experiences with this car driver licensing system. Injury Prevention, 1, 177−181.

Clarke D, Ward P, Truman W & Bartle C (2007). Fatal Vehicle-occupant Collisions: An in-depth Study. Road Safety
Research Report No. 75, Department for Transport, London
Goodwin A H, Wells J K, Foss R D and Williams A F (2006). Encouraging compliance with graduated driver licensing
restrictions. Journal of Safety Research 37 (2006) 343-351.

Knox D, Turner B and Silcock D (2003). Research into Unlicensed Driving. Road Safety Research Report No. 48,
Department for Transport, London

Mayhew D R, Simpson H M, Ferguson S A &Williams A F (1998). Graduated licensing in Nova Scotia: A survey of
teenagers and parents. Journal of Traffic Medicine, 26, 37−44.

Ward H, Christie N, Lyons R, Broughton J, Clarke D & Ward P (2007). Trends in Fatal Car-occupant Accidents, Road
Safety Research Report No. 76, Department for Transport, London

Wells P, Tong S, Sexton B, Grayson G and Jones E (2008). Cohort II: a study of learner and new drivers (Volume 1).
Road Safety Research Report 81, Department for Transport, London.

Williams A F, Nelson L A & Leaf W A (2002). Responses of teenagers and their parents to California's graduated
licensing system. Accident Analysis and Prevention, 34, 835−842.




                                                          A17
Appendix 1. Methodological assumptions
Key assumptions underlying the estimation methodology are:

•   Training received and age of newly qualified driver - In order to obtain estimates for the proposed
    scenarios, it was necessary to make certain assumptions. For example, if a learner has to have a 12-
    month learning period then it was assumed they still take what training they currently feel is necessary to
    pass their test. They will thus have the same chance of passing the driving test and, once they pass, will
    still drive in a similar way. They will, of course, be slightly older and will accumulate their driving experience
    slightly later than is current, but it is not tenable to assume they have a better chance of passing or will be
    a ‘safer’ driver – other than through the slight maturation effect.

•   Demand for test - Where new drivers cannot take their practical test at the same time point as they do
    currently, then it is anticipated there will be an increased demand for tests as soon as these drivers
    become eligible (e.g. where the measure proposed is to raise the age at which they can sit their driving test
    to 18 years). For example, where current 17-year olds cannot take their test until they are 18 years old,
    then there will be a large increase in 18-year olds coming to test early. It was necessary to make an
    assumption on the distribution of 18-year olds passing. In this context different scenarios were considered:

       o   That current 17-year olds come to test when they are aged 18 years in a similar way to how they
           came when they were 17 years (Scenario 1(a) in Section 4.2 of this annex);
       o   They will take their test sooner in their 18th year than when they were able to take the test in their
           17th year , because they will have been preparing for the test while they were still 17 years old
           (Scenario 1, situation (b));

•   Driver population - The 1% sample of DVLA data held by TRL has been used to extract numbers of
    drivers by ‘year of age’: ‘month of age’ in 2006. Tables of numbers of drivers by age and experience were
    constructed to reflect the number of drivers in the current licensing system. The table of drivers (using
    2006 data) includes numbers in the driving population for age (17 years: 00 months to 21years: 11months)
    and experience up to 4 years: 11 months.

•   Accident prediction model - Using this table of counts, a simple KSI accident model can be used to
    predict casualties. These can then be compared with the observed STATS19 KSI casualties (avoiding
    double counting). The initial accident model was based on data from Wells et al., (2008), but needed some
    adjustment to match the observed STATS19 data; that adjustment incorporates an accident to casualty
    factor. The model suggested has the same structural form as that for the ‘best’ generalised linear model
    using data from Wells et al. The model was a reasonable predictor (and was made to fit the 17-year olds
    exactly), see Appendix 3, Table 2.

•   Invariance of accident model - The estimated number of casualties is determined by using the same
    accident prediction model regardless of the scenario being evaluated. The fact that a different distribution
    of drivers is being modelled is what leads to different estimates. The total estimated reductions in KSI
    casualties are computed and may be adjusted to allow for issues such as non-compliance and where the
    driver of interest did not contribute to the cause of the accident.




                                                         A18
Appendix 2. Simulating the population of drivers by age and experience
The basic approach to simulating the population of drivers by age and experience was to develop a table of
the count of drivers in the population, where the columns and rows are respectively the age of the driver (in
years and months) and driving experience (in months). This table was based on DVLA data for 2006 and
does, in effect, simulate the expected numbers of drivers under the various licensing scenarios. Simulating the
driver population was necessary in order that different populations could be generated depending on the
licensing scenario being considered. The actual DVLA database could not be used for this purpose as it only
reflects the driving population for the existing licensing regulations.

Data from a 1% sample of DVLA drivers was extracted in order to derive the numbers of new drivers in 2006
by the age and month of the driver when passing their test. The numbers of drivers of each age as at the end
of 2006 were also extracted. A table of counts of drivers was constructed which estimated how many drivers
by ‘year of age’ and ‘month of age’ had between 1 month and 60 months driving experience. The approach
ensured that the population of all drivers from age 17 years to 21 years was available for estimating casualties.

The construction method used for this table is perhaps best illustrated by looking at an example. It was found
that there were 1,300 drivers aged 17 years 0 months who entered the DVLA register in 2006. If we assume a
uniform distribution across the year for when these 1,300 drivers passed their test, then at any stage there
were 108 (=1300/12) with an average of ½ month experience, 108 with an average of 1½-months experience
(who were aged 17 years 2 months) etc.

Similarly, given that there were, over a whole year, 5,500 drivers aged 17 years 1 month when they passed
their test, there will be 458 (=5500/12) with an average of ½-month experience, 458 with an average of 1½
months experience (who will then be aged 17 years 2 months) etc. In this way Table 1 was constructed:

            Table 1. Table of estimated population numbers of drivers by age and experience

                  Driver age at beginning of month (i.e. 17 years plus months)
  Experience
  (at start of      17yr        17yr       17yr       17yr      17yr       17yr
    month)          0mth        1mth       2mth       3mth      4mth       5mth        …
     0mth            108         458        692       1367      2017       1958        …
     1mth                        108        458        692      1367       2017        …
     2mth                                   108        458       692       1367        …
     3mth                                              108       458        692        …
     4mth                                                        108        458        …
     5mth                                                                   108        …
     6mth                                                                              …
     7mth                                                                              …
     8mth
     9mth
       ….

This process was continued for drivers who passed their test when aged 18, 19, 20 or 21 years. The row
entries correspond to the experience gained (in the first 5 years) and will, for 17-year olds, have included their
first 5 years of driving, but for 18-year olds, just the first 4 years of driving etc. In theory, this process could be
continued to create a whole population table, but it would be very large and would need build in a natural end
to driving careers (i.e. beyond a certain age there are very few new drivers coming into the system and there
are existing drivers who are giving-up driving). In practice, it was felt that including 17 to 21-year old new




                                                          A19
drivers for their first few years of driving provides an appropriate simulation for most of the driver licensing
scenarios being considered (and where this is not the case, the estimated casualties can be scaled up as
appropriate). However, a more extensive table was constructed when estimating the proportion of drivers
with less than two years of driving experience.

An example from part of this table is given below, Table 2, and shows some older 17-year olds and younger
18-year olds. It shows the 108 17-year olds gaining experience month by month and how the table of counts is
growing by age and experience.

            Table 2. Table of estimated population numbers of drivers by age and experience

                                       Driver age at beginning of month
  Experience
  (at start of            17yr       17yr      18yr      18yr       18yr       18yr
    month)        …      10mth      11mth      0mth      1mth       2mth       3mth        …
     0mth         …       1850       1800      1658      1425       1367       1200        …
     1mth         …       1800       1850      1800      1658       1425       1367        …
     2mth         …       2017       1800      1850      1800       1658       1425        …
     3mth         …       1842       2017      1800      1850       1800       1658        …
     4mth         …       1958       1842      2017      1800       1850       1800        …
     5mth         …       2017       1958      1842      2017       1800       1850        …
     6mth         …       1658       2017      1958      1842       2017       1800        …
     7mth         …       1367       1658      2017      1958       1842       2017        …
     8mth         …        692       1367      1658      2017       1958       1842        …
     9mth         …        458        692      1367      1658       2017       1958        …
     10mth                 108        458       692      1367       1658       2017        …
     11mth                            108       458       692       1367       1658        …
   1 yr 0mth                                    108       458        692       1367        …
   1 yr 1mth                                              108        458        692        …
   1 yr 2mth                                                         108        458        …
   1 yr 3mth                                                                    108        …
   1 yr 4mth                                                                               …

Hence, at any time, the number of drivers aged, for example, 18 years 0mths was estimated as the sum of the
numbers in that column in Table 2 (that is, 1658 + 1800 + 1850 + …. + 108). Applying this approach for all
ages from 17 years to 21 years and 11 months, the total expected population numbers for 17-year olds
through to 21-year olds can be obtained. These are shown in Table A.1, where it can be seen that, whilst
these do not exactly match the DVLA figures (due to year to year fluctuations), nevertheless the agreement
between the simulated and the observed numbers is quite close.




                                                        A20
Appendix 3. Accident / casualty liability models
Data from the Cohort II study (Wells et al., 2008) has been used to develop accident liability models. The age
and experience of the driver is known and the severity of any accident was reported by the survey respondent
in that study. These data were used in the initial development of accident liability models. These models did
not include any exposure variables because it was assumed that drivers will continue to drive similar miles and
frequency as currently and, as such, the exposure is subsumed within the age and experience effect (also
exposure data are not readily available). The analysis on those injury accidents where there was a severe or a
fatal casualty suggested the following functional equation provided the ‘best’ fitting relationship (Equation 1):

                      accident liability (pa) = e(A) e(B / age) e(C / experience)             ………… (1)

       where
       ‘age’ is that at the mid-point of the reporting period when the accident occurred 78
       ‘experience’ is that at the mid-point of the reporting period

This structural form gives a stronger effect for small values of experience (i.e. the impact on accident liability is
stronger for lower levels of experience as compared to higher levels of experience). The model derived using
data from the Cohort II study predicts accident liability, but it is casualties which are of interest when
investigating the identified scenarios. Hence, this model was adjusted to reflect the casualty-to-accident ratio
and was adjusted to fit as closely as possible to the observed KSI casualty numbers from STATS19 data.

STATS19 data consists of all injury accidents recorded by the police. However, it is widely recognised that
there is a degree of under-reporting so the database does not include all slight injury accidents but will
include virtually all accidents where there was a fatality and most where a severe injury was received. An
accident may involve one or more vehicles and the database includes details of all involved vehicles and
associated casualties. The primary interest is in younger drivers and the number of KSI casualties (of any age)
from accidents where a younger person was driving. This is simple to calculate for single vehicle accidents,
because there was only one driver. However, for accidents involving two or more vehicles it is more complex
because either driver could be, for example, aged 17-years old.

A table for those accidents involving two vehicles was constructed such that the actual number of casualties
could be estimated where one and/or both vehicles were driven by a 17-year old. In this way double-counting
could be properly accounted for. The adjustment to casualty numbers to avoid double-counting in two-vehicle
accidents was almost seven per cent for 17-year olds. This figure was derived by comparing the numbers of
casualties in two-vehicle accidents when taking into account double counting compared with not taking it into
account. Given that over 60% of accidents involve two vehicles this is a substantial reduction.

The situation for accidents involving three or more vehicles is even more complex. However, these only
contribute about 12% of all car accidents and, assuming the level of adjustment to allow for double counting is
similar to that for two-vehicle accidents, then the double (or triple) counting effect cannot be that great. Hence,
the observed numbers of KSI casualties found in the 2006 STATS19 database were derived by taking those
from single vehicle accidents, those from two-vehicle accidents adjusted to exclude double counting and those
from three or more vehicle accidents where there is an element of double (or triple) counting. The table of KSI
casualties thus derived for drivers aged 17 to 21 years, together with the KSI liability rate, is as shown in Table
1 below.

            78
               New drivers in the Cohort II study reported on their driving experience at six months, 12-months, 24-months and 36-months, hence
            the age of the driver when they had an accident in the first 6-months of driving would be the age at which they passed plus 3-months,
            i.e. half the length of reporting period.




                                                                      A21
            Table 1 Observed number of KSI casualties involving young drivers (STATS19 for 2006)

                                              KSI casualties
                     Driver       Single     Two-                              DVLA         KSI
                      age         vehicle   vehicle 2+vehicles      Total     drivers      rate
                       17           369       382        120         871      78,700      0.0111
                       18           561       639        277        1477      292,700     0.0050
                       19           540       562        256        1358      385,000     0.0035
                       20           471       539        216        1226      422,100     0.0029
                       21           403       517        184        1104      452,700     0.0024

The accident liability model discussed above was adjusted to reflect the numbers in the above table, taking
into account the simulated population of drivers as given by Table A.1. The functional form of the resultant
model included an extra constant term which helped when estimating the experience effect (Equation 2):

                   KSI accident liability (pa) = e(A) e(B / age) {e(C / experience) + D} ………… (2)

As stated above, the parameter estimates were adjusted to allow for the accident-to-KSI casualty ratio to be
incorporated such that estimates from the simulation closely matched the observed data. The predicted and
observed KSI casualties are shown in Table 2, together with the percentage difference of predicted compared
to observed KSI casualties. There is a reasonable agreement between the observed and fitted values for
drivers aged from 17 years to 20 years.

            Table 2 Observed and predicted KSI casualties involving young drivers

                                      Observed
                                        KSI
                                     casualties   Predicted
                                     (STATS19,       KSI             %      Average      Average
                     Driver age         2006)     casualties   difference     age       experience
                         17             871          871           0.0%      17.72        0.283
                         18             1477        1527           3.4%      18.55        0.722
                         19             1358        1217         -10.4%      19.51        1.453
                         20             1226        1179          -3.8%      20.51        2.290
                         21             1104        1168           5.8%      21.51        3.106

There is perhaps some tendency for the STATS19 observed (attributed) casualties to be underestimated by
the predicted values, especially for the 19-year olds. This discrepancy decreases and then appears to reverse,
such that the casualty numbers are over-estimated. It is important to remember that the accident liability
function has been set to exactly estimate the 17-year olds, which is justified because these are the main age
group affected by the proposed scenarios.

The prediction equation with estimated parameters used for the above was (Equation 3):

                KSI accident liability (pa) = e(-7.2) e(23 / age) {e(0.11 / experience) + 0.06} ………. (3)
This was the ‘best’ fitting model for KSI casualties. However, it is not perfect and as such any results using this
equation and the simulated population of drivers will have some associated ‘noise’.
associated ‘noise’.




                                                        A22
      Annex B

      Estimated costs and benefits of options - best and worst case scenarios

      Option B - Best Case Scenario (Minimum Costs, Maximum Benefits)

      Year                                               1                 2             3             4             5            6              7            8              9            10             Total
      Discount Rate = 3.5%                               1              0.9662        0.9335        0.9019        0.8714        0.842         0.8135        0.786         0.7594        0.7337


      Costs
      Pre-driver qualification set up                    2,400,000        2,400,000
      Pre-driver educational training                   35,000,000       35,000,000    35,000,000    35,000,000    35,000,000    35,000,000    35,000,000    35,000,000    35,000,000    35,000,000
      Retraining ADIs                                    2,150,000        2,150,000
      Attitude Indicator                                 1,000,000
      Peer Group Sessions                                1,875,000        1,875,000     1,875,000     1,875,000     1,875,000     1,875,000     1,875,000     1,875,000     1,875,000     1,875,000
      Workbook                                           3,750,000        3,750,000     3,750,000     3,750,000     3,750,000     3,750,000     3,750,000     3,750,000     3,750,000     3,750,000
      Theory Test (case studies)                         1,800,000
      New Hazard Perception Test                         2,000,000
      Developing New Practical Test                      2,100,000
      Training Driving Examiners                         1,850,000
      DSA reorganisation                                 6,100,000        6,100,000
      Additional DSA running costs                      15,100,000       15,100,000    15,100,000    15,100,000    15,100,000    15,100,000    15,100,000    15,100,000    15,100,000    15,100,000

      Total Costs                                       75,125,000       66,375,000    55,725,000    55,725,000    55,725,000    55,725,000    55,725,000    55,725,000    55,725,000    55,725,000
      Total PV Costs                                    75,125,000       64,131,525    52,019,288    50,258,378    48,558,765    46,920,450    45,332,288    43,799,850    42,317,565    40,885,433    509,348,540
A23




      Benefits
      Reduced casulties                               252,651,801       252,651,801   252,651,801   252,651,801   252,651,801   252,651,801   252,651,801   252,651,801   252,651,801   252,651,801
      Reduced damage only accidents                   338,730,440       338,730,440   338,730,440   338,730,440   338,730,440   338,730,440   338,730,440   338,730,440   338,730,440   338,730,440
      Faster Learning                                  51,187,500        51,187,500    51,187,500    51,187,500    51,187,500    51,187,500    51,187,500    51,187,500    51,187,500    51,187,500
      Higher Pass Rate Structured                       5,729,062         5,729,062     5,729,062     5,729,062     5,729,062     5,729,062     5,729,062     5,729,062     5,729,062     5,729,062
      Higher Pass Rate (S) Lesson savings              14,883,750        14,883,750    14,883,750    14,883,750    14,883,750    14,883,750    14,883,750    14,883,750    14,883,750    14,883,750
      Non structured learning pass rate                 6,001,875         6,001,875     6,001,875     6,001,875     6,001,875     6,001,875     6,001,875     6,001,875     6,001,875     6,001,875
      Non structured learning lessons savings          15,592,500        15,592,500    15,592,500    15,592,500    15,592,500    15,592,500    15,592,500    15,592,500    15,592,500    15,592,500
      Decoupling Theory Test                            5,523,750         5,523,750     5,523,750     5,523,750     5,523,750     5,523,750     5,523,750     5,523,750     5,523,750     5,523,750
      Test Feedback                                    59,157,500        59,157,500    59,157,500    59,157,500    59,157,500    59,157,500    59,157,500    59,157,500    59,157,500    59,157,500

      Total Benefits                                  749,458,178       749,458,178   749,458,178   749,458,178   749,458,178   749,458,178   749,458,178   749,458,178   749,458,178   749,458,178
      Total PV Benefits                               749,458,178       724,126,492   699,619,209   675,936,331   653,077,856   631,043,786   609,684,228   589,074,128   569,138,541   549,877,465   6,451,036,214
      Total Net Present Value                         674,333,178       659,994,967   647,599,922   625,677,953   604,519,091   584,123,336   564,351,940   545,274,278   526,820,976   508,992,033   5,941,687,674
      Option B - Worst Case Scenario (Maximum Costs, Minimum Benefits)

      Year                                               1                  2             3             4             5            6              7            8              9            10             Total
      Discount Rate = 3.5%                               1               0.9662        0.9335        0.9019        0.8714        0.842         0.8135        0.786         0.7594        0.7337


      Costs
      Pre-driver qualification set up                   3,000,000          3,000,000
      Pre-driver educational training                  35,000,000         35,000,000    35,000,000    35,000,000    35,000,000    35,000,000    35,000,000    35,000,000    35,000,000    35,000,000
      Retraining ADIs                                   5,350,000          5,350,000
      Attitude Indicator                                1,000,000
      Peer Group Sessions                               1,875,000          1,875,000     1,875,000     1,875,000     1,875,000     1,875,000     1,875,000     1,875,000     1,875,000     1,875,000
      Workbook                                          3,750,000          3,750,000     3,750,000     3,750,000     3,750,000     3,750,000     3,750,000     3,750,000     3,750,000     3,750,000
      Theory Test (case studies)                        1,800,000
      New Hazard Perception Test                        2,000,000
      Developing New Practical Test                     5,700,000
      Training Driving Examiners                        5,550,000
      DSA reorganisation                               10,650,000         10,650,000
      Additional DSA running costs                     15,100,000         15,100,000    15,100,000    15,100,000    15,100,000    15,100,000    15,100,000    15,100,000    15,100,000    15,100,000

      Total Costs                                      90,775,000         74,725,000    55,725,000    55,725,000    55,725,000    55,725,000    55,725,000    55,725,000    55,725,000    55,725,000
      Total PV Costs                                   90,775,000         72,199,295    52,019,288    50,258,378    48,558,765    46,920,450    45,332,288    43,799,850    42,317,565    40,885,433    533,066,310


      Benefits
      Reduced casulties                                63,162,950         63,162,950    63,162,950    63,162,950    63,162,950    63,162,950    63,162,950    63,162,950    63,162,950    63,162,950
      Reduced damage only accidents                    84,682,610         84,682,610    84,682,610    84,682,610    84,682,610    84,682,610    84,682,610    84,682,610    84,682,610    84,682,610
      Faster Learning                                  51,187,500         51,187,500    51,187,500    51,187,500    51,187,500    51,187,500    51,187,500    51,187,500    51,187,500    51,187,500
      Higher Pass Rate Structured                       5,729,062          5,729,062     5,729,062     5,729,062     5,729,062     5,729,062     5,729,062     5,729,062     5,729,062     5,729,062
A24




      Higher Pass Rate (S) Lesson savings              14,883,750         14,883,750    14,883,750    14,883,750    14,883,750    14,883,750    14,883,750    14,883,750    14,883,750    14,883,750
      Non structured learning pass rate                 6,001,875          6,001,875     6,001,875     6,001,875     6,001,875     6,001,875     6,001,875     6,001,875     6,001,875     6,001,875
      Non structured learning lessons savings          15,592,500         15,592,500    15,592,500    15,592,500    15,592,500    15,592,500    15,592,500    15,592,500    15,592,500    15,592,500
      Decoupling Theory Test                            5,523,750          5,523,750     5,523,750     5,523,750     5,523,750     5,523,750     5,523,750     5,523,750     5,523,750     5,523,750
      Test Feedback                                    46,305,000         46,305,000    46,305,000    46,305,000    46,305,000    46,305,000    46,305,000    46,305,000    46,305,000    46,305,000

      Total Benefits                                  293,068,997        293,068,997   293,068,997   293,068,997   293,068,997   293,068,997   293,068,997   293,068,997   293,068,997   293,068,997
      Total PV Benefits                               293,068,997        283,163,265   273,579,909   264,318,929   255,380,324   246,764,096   238,411,629   230,352,232   222,556,597   215,024,723   2,522,620,701
      Total Net Present Value                                                                                                                                                                          1,989,554,391
      Option C - Best Case Scenario (Minimum Costs, Maximum Benefits)

      Year                                               1                  2              3              4               5             6               7             8              9             10                   Total
      Discount Rate = 3.5%                               1               0.9662         0.9335         0.9019          0.8714         0.842          0.8135         0.786         0.7594         0.7337


      Costs
      Minimum Learning Period                           44,887,500        44,887,500     44,887,500     44,887,500     44,887,500     44,887,500     44,887,500     44,887,500     44,887,500     44,887,500
      Full Driving Age                                  82,500,000        82,500,000     82,500,000     82,500,000     82,500,000     82,500,000     82,500,000     82,500,000     82,500,000     82,500,000
      Minimum Number of Lessons                          3,307,500         3,307,500      3,307,500      3,307,500      3,307,500      3,307,500      3,307,500      3,307,500      3,307,500      3,307,500
      Reducing Provisional age -casualty accidents      60,672,000        60,672,000     60,672,000     60,672,000     60,672,000     60,672,000     60,672,000     60,672,000     60,672,000     60,672,000
      Reducing Provisional age - damage only            41,474,000        41,474,000     41,474,000     41,474,000     41,474,000     41,474,000     41,474,000     41,474,000     41,474,000     41,474,000

      Total Costs                                     232,841,000        232,841,000    232,841,000    232,841,000    232,841,000    232,841,000    232,841,000    232,841,000    232,841,000    232,841,000
      Total PV Costs                                  232,841,000        224,970,974    217,357,074    209,999,298    202,897,647    196,052,122    189,416,154    183,013,026    176,819,455    170,835,442         2,004,202,192


      Benefits
      Reduced casulties                                 55,635,759        55,635,759     55,635,759     55,635,759     55,635,759     55,635,759     55,635,759     55,635,759     55,635,759     55,635,759
      Reduced damage only accidents                     37,295,450        37,295,450     37,295,450     37,295,450     37,295,450     37,295,450     37,295,450     37,295,450     37,295,450     37,295,450

      Total Benefits                                    92,931,209        92,931,209     92,931,209     92,931,209     92,931,209     92,931,209     92,931,209     92,931,209     92,931,209     92,931,209
      Total PV Benefits                                 92,931,209        89,790,134     86,751,284     83,814,658     80,980,256     78,248,078     75,599,539     73,043,931     70,571,960     68,183,628           799,914,677
      NPV                                             -139,909,791       -135,180,840   -130,605,790   -126,184,640   -121,917,392   -117,804,044   -113,816,615   -109,969,095   -106,247,495   -102,651,813        -1,204,287,514



      Option C - Worst Case Scenario (Maximum Costs, Minimum Benefits)
A25




      Year                                               1                  2              3              4              5              6               7             8              9             10                   Total
      Discount Rate = 3.5%                               1               0.9662         0.9335         0.9019         0.8714          0.842          0.8135         0.786         0.7594         0.7337


      Costs                                                                                                                                                                                                     PV
      Minimum Learning Period                          84,375,000         84,375,000     84,375,000     84,375,000     84,375,000     84,375,000     84,375,000     84,375,000     84,375,000     84,375,000
      Full Driving Age                                154,687,500        154,687,500    154,687,500    154,687,500    154,687,500    154,687,500    154,687,500    154,687,500    154,687,500    154,687,500
      Minimum Number of Lessons                        94,500,000         94,500,000     94,500,000     94,500,000     94,500,000     94,500,000     94,500,000     94,500,000     94,500,000     94,500,000
      Reducing Provisional age -casualty accidents     60,672,000         60,672,000     60,672,000     60,672,000     60,672,000     60,672,000     60,672,000     60,672,000     60,672,000     60,672,000
      Reducing Provisional age - damage only           41,474,000         41,474,000     41,474,000     41,474,000     41,474,000     41,474,000     41,474,000     41,474,000     41,474,000     41,474,000

      Total Costs                                     435,708,500        435,708,500    435,708,500    435,708,500    435,708,500    435,708,500    435,708,500    435,708,500    435,708,500    435,708,500
      Total PV Costs                                  435,708,500        420,981,553    406,733,885    392,965,496    379,676,387    366,866,557    354,448,865    342,466,881    330,877,035    319,679,326         3,750,404,485


      Benefits
      Reduced casulties                                 11,127,152        11,127,152     11,127,152     11,127,152     11,127,152     11,127,152     11,127,152     11,127,152     11,127,152     11,127,152
      Reduced damage only accidents                      7,459,090         7,459,090      7,459,090      7,459,090      7,459,090      7,459,090      7,459,090      7,459,090      7,459,090      7,459,090

      Total Benefits                                    18,586,242        18,586,242     18,586,242     18,586,242     18,586,242     18,586,242     18,586,242     18,586,242     18,586,242     18,586,242
      Total PV Benefits                                 18,586,242        17,958,027     17,350,257     16,762,932     16,196,051     15,649,616     15,119,908     14,608,786     14,114,392     13,636,726           159,982,935

      NPV                                             -417,122,258       -403,023,526   -389,383,628   -376,202,565   -363,480,336   -351,216,941   -339,328,957   -327,858,095   -316,762,643   -306,042,601        -3,590,421,549
      Option D - Best Case Scenario (Minimum Costs, Maximum Benefits)

      Year                                               1                  2              3              4              5              6              7              8              9             10              Total
      Discount Rate = 3.5%                               1               0.9662         0.9335         0.9019         0.8714          0.842         0.8135          0.786         0.7594         0.7337


      Costs
      Restricted Licence                              300,000,000        300,000,000    300,000,000    300,000,000    300,000,000    300,000,000    300,000,000    300,000,000    300,000,000    300,000,000

      Total Costs                                     300,000,000        300,000,000    300,000,000    300,000,000    300,000,000    300,000,000    300,000,000    300,000,000    300,000,000    300,000,000
      Total PV Costs                                  300,000,000        289,860,000    280,050,000    270,570,000    261,420,000    252,600,000    244,050,000    235,800,000    227,820,000    220,110,000    2,582,280,000


      Benefits
      Reduced casulties                               130,907,669        130,907,669    130,907,669    130,907,669    130,907,669    130,907,669    130,907,669    130,907,669    130,907,669    130,907,669
      Reduced damage only accidents                    87,754,000         87,754,000     87,754,000     87,754,000     87,754,000     87,754,000     87,754,000     87,754,000     87,754,000     87,754,000

      Total Benefits                                  218,661,669        218,661,669    218,661,669    218,661,669    218,661,669    218,661,669    218,661,669    218,661,669    218,661,669    218,661,669
      Total PV Benefits                               218,661,669        211,270,905    204,120,668    197,210,959    190,541,778    184,113,125    177,881,268    171,868,072    166,051,671    160,432,067    1,882,152,182
      NPV                                              -81,338,331        -78,589,095    -75,929,332    -73,359,041    -70,878,222    -68,486,875    -66,168,732    -63,931,928    -61,768,329    -59,677,933    -700,127,818



      Option D - Worst Case Scenario (Maximum Costs, Minimum Benefits)

      Year                                               1                  2              3              4              5              6              7              8              9             10              Total
      Discount Rate = 3.5%                               1               0.9662         0.9335         0.9019         0.8714          0.842         0.8135          0.786         0.7594         0.7337
A26




      Costs
      Restricted Licence                              562,500,000        562,500,000    562,500,000    562,500,000    562,500,000    562,500,000    562,500,000    562,500,000    562,500,000    562,500,000

      Total Costs                                     562,500,000        562,500,000    562,500,000    562,500,000    562,500,000    562,500,000    562,500,000    562,500,000    562,500,000    562,500,000
      Total PV Costs                                  562,500,000        543,487,500    525,093,750    507,318,750    490,162,500    473,625,000    457,593,750    442,125,000    427,162,500    412,706,250    4,841,775,000


      Benefits
      Reduced casulties                                 65,453,835        65,453,835     65,453,835     65,453,835     65,453,835     65,453,835     65,453,835     65,453,835     65,453,835     65,453,835
      Reduced damage only accidents                     43,877,000        43,877,000     43,877,000     43,877,000     43,877,000     43,877,000     43,877,000     43,877,000     43,877,000     43,877,000

      Total Benefits                                  109,330,835        109,330,835    109,330,835    109,330,835    109,330,835    109,330,835    109,330,835    109,330,835    109,330,835    109,330,835
      Total PV Benefits                               109,330,835        105,635,452    102,060,334     98,605,480     95,270,889     92,056,563     88,940,634     85,934,036     83,025,836     80,216,033      941,076,091

      NPV                                             -453,169,166       -437,852,048   -423,033,416   -408,713,270   -394,891,611   -381,568,437   -368,653,116   -356,190,964   -344,136,664   -332,490,217   -3,900,698,909
      Option E - Best Case Scenario (Minimum Costs, Maximum Benefits)

      Year                                               1                 2             3             4             5            6              7            8              9            10            Total
      Discount Rate = 3.5%                               1              0.9662        0.9335        0.9019        0.8714        0.842         0.8135        0.786         0.7594        0.7337


      Costs
      Restricted Licence                              300,000,000       300,000,000   300,000,000   300,000,000   300,000,000   300,000,000   300,000,000   300,000,000   300,000,000   300,000,000
      Pre-driver qualification set up                   2,400,000         2,400,000
      Pre-driver educational training                  35,000,000        35,000,000    35,000,000    35,000,000    35,000,000    35,000,000    35,000,000    35,000,000    35,000,000    35,000,000
      Retraining ADIs                                   2,150,000         2,150,000
      Attitude Indicator                                1,000,000
      Peer Group Sessions                               1,875,000         1,875,000     1,875,000     1,875,000     1,875,000     1,875,000     1,875,000     1,875,000     1,875,000     1,875,000
      Workbook                                          3,750,000         3,750,000     3,750,000     3,750,000     3,750,000     3,750,000     3,750,000     3,750,000     3,750,000     3,750,000
      Theory Test (case studies)                        1,800,000
      New Hazard Perception Test                        2,000,000
      Developing New Practical Test                     2,100,000
      Training Driving Examiners                        1,850,000
      DSA reorganisation                                6,100,000         6,100,000
      Additional DSA running costs                     15,100,000        15,100,000    15,100,000    15,100,000    15,100,000    15,100,000    15,100,000    15,100,000    15,100,000    15,100,000

      Total Costs                                     375,125,000       366,375,000   355,725,000   355,725,000   355,725,000   355,725,000   355,725,000   355,725,000   355,725,000   355,725,000
      Total PV Costs                                  375,125,000       353,991,525   332,069,288   320,828,378   309,978,765   299,520,450   289,382,288   279,599,850   270,137,565   260,995,433   3,091,628,540


      Benefits
      Reduced casulties                               243,488,264       243,488,264   243,488,264   243,488,264   243,488,264   243,488,264   243,488,264   243,488,264   243,488,264   243,488,264
      Reduced damage only accidents                   163,222,440       163,222,440   163,222,440   163,222,440   163,222,440   163,222,440   163,222,440   163,222,440   163,222,440   163,222,440
      Faster Learning                                  51,187,500        51,187,500    51,187,500    51,187,500    51,187,500    51,187,500    51,187,500    51,187,500    51,187,500    51,187,500
A27




      Higher Pass Rate Structured                       5,729,062         5,729,062     5,729,062     5,729,062     5,729,062     5,729,062     5,729,062     5,729,062     5,729,062     5,729,062
      Higher Pass Rate (S) Lesson savings              14,883,750        14,883,750    14,883,750    14,883,750    14,883,750    14,883,750    14,883,750    14,883,750    14,883,750    14,883,750
      Non structured learning pass rate                 6,001,875         6,001,875     6,001,875     6,001,875     6,001,875     6,001,875     6,001,875     6,001,875     6,001,875     6,001,875
      Non structured learning lessons savings          15,592,500        15,592,500    15,592,500    15,592,500    15,592,500    15,592,500    15,592,500    15,592,500    15,592,500    15,592,500
      Decoupling Theory Test                            5,523,750         5,523,750     5,523,750     5,523,750     5,523,750     5,523,750     5,523,750     5,523,750     5,523,750     5,523,750
      Test Feedback                                    69,300,000        69,300,000    69,300,000    69,300,000    69,300,000    69,300,000    69,300,000    69,300,000    69,300,000    69,300,000

      Total Benefits                                  574,929,141       574,929,141   574,929,141   574,929,141   574,929,141   574,929,141   574,929,141   574,929,141   574,929,141   574,929,141
      Total PV Benefits                               574,929,141       555,496,536   536,696,353   518,528,593   500,993,254   484,090,337   467,704,856   451,894,305   436,601,190   421,825,511   4,948,760,077
      NPV                                             199,804,141       201,505,011   204,627,066   197,700,215   191,014,489   184,569,887   178,322,569   172,294,455   166,463,625   160,830,079   1,857,131,537
      Option E - Worst Case Scenario (Maximum Costs, Minimum Benefits)

      Year                                               1                  2              3              4              5              6              7              8              9             10              Total
      Discount Rate = 3.5%                               1               0.9662         0.9335         0.9019         0.8714          0.842         0.8135          0.786         0.7594         0.7337


      Costs
      Restricted Licence                              562,500,000        562,500,000    562,500,000    562,500,000    562,500,000    562,500,000    562,500,000    562,500,000    562,500,000    562,500,000
      Pre-driver qualification set up                   3,000,000          3,000,000
      Pre-driver educational training                  35,000,000         35,000,000     35,000,000     35,000,000     35,000,000     35,000,000     35,000,000     35,000,000     35,000,000     35,000,000
      Retraining ADIs                                   5,350,000          5,350,000
      Attitude Indicator                                1,000,000
      Peer Group Sessions                               1,875,000          1,875,000      1,875,000      1,875,000      1,875,000      1,875,000      1,875,000      1,875,000      1,875,000      1,875,000
      Workbook                                          3,750,000          3,750,000      3,750,000      3,750,000      3,750,000      3,750,000      3,750,000      3,750,000      3,750,000      3,750,000
      Theory Test (case studies)                        1,800,000
      New Hazard Perception Test                        2,000,000
      Developing New Practical Test                     5,700,000
      Training Driving Examiners                        5,550,000
      DSA reorganisation                               10,650,000         10,650,000
      Additional DSA running costs                     15,100,000         15,100,000     15,100,000     15,100,000     15,100,000     15,100,000     15,100,000     15,100,000     15,100,000     15,100,000

      Total Costs                                     653,275,000        637,225,000    618,225,000    618,225,000    618,225,000    618,225,000    618,225,000    618,225,000    618,225,000    618,225,000
      Total PV Costs                                  653,275,000        615,686,795    577,113,038    557,577,128    538,721,265    520,545,450    502,926,038    485,924,850    469,480,065    453,591,683    5,374,841,310


      Benefits
      Reduced casulties                                91,308,099         91,308,099     91,308,099     91,308,099     91,308,099     91,308,099     91,308,099     91,308,099     91,308,099     91,308,099
      Reduced damage only accidents                    61,208,415         61,208,415     61,208,415     61,208,415     61,208,415     61,208,415     61,208,415     61,208,415     61,208,415     61,208,415
      Faster Learning                                  51,187,500         51,187,500     51,187,500     51,187,500     51,187,500     51,187,500     51,187,500     51,187,500     51,187,500     51,187,500
A28




      Higher Pass Rate Structured                       5,729,062          5,729,062      5,729,062      5,729,062      5,729,062      5,729,062      5,729,062      5,729,062      5,729,062      5,729,062
      Higher Pass Rate (S) Lesson savings              14,883,750         14,883,750     14,883,750     14,883,750     14,883,750     14,883,750     14,883,750     14,883,750     14,883,750     14,883,750
      Non structured learning pass rate                 6,001,875          6,001,875      6,001,875      6,001,875      6,001,875      6,001,875      6,001,875      6,001,875      6,001,875      6,001,875
      Non structured learning lessons savings          15,592,500         15,592,500     15,592,500     15,592,500     15,592,500     15,592,500     15,592,500     15,592,500     15,592,500     15,592,500
      Decoupling Theory Test                            5,523,750          5,523,750      5,523,750      5,523,750      5,523,750      5,523,750      5,523,750      5,523,750      5,523,750      5,523,750
      Test Feedback                                    46,200,000         46,200,000     46,200,000     46,200,000     46,200,000     46,200,000     46,200,000     46,200,000     46,200,000     46,200,000

      Total Costs                                     297,634,951        297,634,951    297,634,951    297,634,951    297,634,951    297,634,951    297,634,951    297,634,951    297,634,951    297,634,951
      Total PV Benefits                               297,634,951        287,574,890    277,842,227    268,436,962    259,359,096    250,608,629    242,126,033    233,941,072    226,023,982    218,374,764    2,561,922,605

      NPV                                             -355,640,049       -328,111,905   -299,270,811   -289,140,165   -279,362,169   -269,936,821   -260,800,005   -251,983,778   -243,456,083   -235,216,919   -2,812,918,705
1   ECOTEC

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:6
posted:7/1/2011
language:English
pages:71