Transport Ten Year Plan 2000: background analysis

Introduction

This paper provides an overview of the modelling and analytical work that has informed the Government's 10 Year Plan for transport: Transport 2010: The 10 Year Plan [1].

In developing the Plan we have focused on the broad outcomes that different investment and policy options can deliver over the next ten years, and their cost-effectiveness, within the overall framework of promoting sustainable development. Our modelling work to support this analysis has had a correspondingly broad focus. It has involved:

  • identifying a baseline of what would happen to key transport outcomes in the absence of the measures set out in the Plan. This has illustrated the challenges that we face. Chapter 1 sets out the key influences on transport demand and the core assumptions we have made. Chapter 2 describes the models we have used and the baseline assessment;
  • estimating the overall impact of the Plan and looking at the cost-effectiveness of different components of the Plan in tackling two of the key challenges: road congestion, and carbon dioxide emissions. We also looked at three illustrative scenarios showing the potential impact of some future changes and policy choices. This work is described in Chapter 3; and
  • considering evidence from the cost-effectiveness analysis, individual project appraisals and other information, identifying in broad terms how the Plan contributes to the key dimensions of sustainable development. This work is described in Chapter 4.

We have substantially developed our strategic modelling capabilities. Even though we do not currently have a fully integrated transport model, we have sought to take into account the links between different modes of transport in order to ensure an integrated approach.

We think our modelling and analysis give a good picture of what will happen with and without the measures in the Plan. But it is important to recognise the limitations of strategic modelling and analysis. A significant range of uncertainty attaches to all model forecasts. More confidence can be placed in the relative forecasts for different scenarios than in their absolute levels. The forecasts derive from a large number of inputs which cannot, in many cases, be predicted with any great accuracy. We have, for example, had to make assumptions about how the funds being made available in the Plan will be spent, when in practice detailed decisions will be made by a large number of different organisations, such as local authorities, public transport operators and the Highways Agency. And our modelling will not have taken into account every possible influence on future travel demand. Nor are we able to model all the outcomes that are of interest. We have used sensitivity tests to examine how some of these uncertainties affect the outcomes we are able to model.

The results of our analysis in some cases identify difficult trade-offs: a particular measure may not have positive effects on all the outcomes we are seeking to improve. The Plan reflects the Government's overall judgement on a package of measures.

Chapter 1:
Key influences and core assumptions

1.Transport decisions by businesses, individuals and households reflect a wide range of social and economic factors. Our analysis of the demand for travel and choice of mode seeks to take account of these factors. We briefly describe here the key influences and the core assumptions that we have used throughout our analysis.

2.The levels of economic activity in the economy and household incomes are key determinants of the demand for travel. Although the relationships change over time, we expect that economic growth will increase demand for business and personal travel as well as increase the requirement for freight movement. In line with the latest HM Treasury assumptions, the Plan assumes that the economy will grow at 2.5% per annum between 2000 and 2005, and by 2.25% per annum between 2005 and 2010[2].

3.Demographic factors are also important. Population increases, growth in the number of households and their location affect the number of journeys made. We have used the latest Government projections for population growth (0.3% pa), household growth (0.7% pa) and household composition[3].

4.Decisions about the amount of travel and the choice of mode are also affected by the relative costs of different means of transport. These costs include both money costs (for example fuel and fares) and the costs associated with, for example, travelling time, unreliability and overcrowding. The measures in the Plan influence many of these costs. But there are other costs that we cannot directly influence and on which we have had to make assumptions. These include the world price of oil. The industry consensus is that the long run sustainable price lies between $14 and $18 per barrel at 1999 prices [4]. We have assumed for our modelling that the price of oil falls from $28 in 2000 to $16 in 2010 (at 1999 prices).

Chapter 2:
The models used and analysis of the challenges

5.We have used a number of models to analyse what would happen to key transport and environmental outcomes[5] in the absence of the Plan. This has created a baseline illustrating the challenges we face, against which we have then assessed the impact of individual components of the Plan, and the Plan as a whole. As Transport 2010: The 10 Year Plan explains, these challenges are in part connected with success economic growth and increased prosperity.

6.The models we have used are:

  • a new model of passenger rail demand prepared by the Department;
  • recent modelling work for the shadow Strategic Rail Authority (sSRA) on the future demand for rail freight;
  • the London Transportation Studies (LTS) model;
  • the National Road Traffic Forecasts (NRTF) framework of models;
  • new models of rail emissions prepared by the Department; and
  • models of concentrations of key local air pollutants developed from those used to inform the Government's Air Quality Strategy for England, Scotland, Wales and Northern Ireland ('Air Quality Strategy').[6]

7.This chapter briefly describes the models, and how they have been brought together to produce baseline forecasts. Figure 1 below summarises the modelling framework and key interactions and influences.

Figure 1: The modelling framework, key linkages, interactions and outputs

Figure 1: The modelling framework, key linkages, interactions and outputs

Modelling passenger rail demand

8.The growth in passenger rail demand in Great Britain since the mid-1990s has been unprecedented, both in terms of length (five years with no sign yet of any slowing down) and rate (over 30% in five years, with growth averaging 6.3% a year over the last four years). Recent growth has been spread across all market segments, but has been particularly strong for Intercity and London commuting travel.

9.We have carried out econometric analysis of total passenger demand and non-commuting demand to establish whether the strong growth in recent years represents a structural shift in favour of passenger rail, or a temporary phenomenon due, for example, to one-off effects of privatisation or delayed recovery from the last recession. We now have a reasonably robust explanation for trends over the past 20 years and one that suggests a continuation of recent growth into the future. We have considered non-commuting and commuting demand separately.

Non-commuting demand

10.The recent rapid growth in non-commuting demand is largely explained by:

  • strong economic growth;
  • a change in the relationship between road traffic growth and economic growth: the ratio has fallen compared with previous periods of strong economic growth, reflecting lower car ownership growth and rising congestion (see paragraph 31 below); and
  • average rail fares falling slightly in real terms, in contrast to real increases over the previous decade.

11. Annex A reports the results of our analysis in more detail. Using this model we have forecast future non-commuting demand over the next decade making the following assumptions (in addition to the GDP growth assumption set out in Chapter 1):

  • Fares. The current policy for regulated fares (about half the total) continues throughout the Plan period: they change by 1% below the retail price index (RPI). Unregulated fares change in line with RPI; and
  • Car traffic. We have used the baseline forecast of traffic growth produced by the NRTF modelling framework. This averages 2% per annum, and continues the recent loosening of the link between economic growth and overall car traffic growth acting to increase rail demand.

12.We estimate that these factors alone will increase non-commuting demand by 39% over the Plan period.

Commuting demand

13.For commuting demand dominated by commuting to central London we have assumed that the rate of growth slows down from recent high rates, with an overall assumed increase of 15% between 2000/01 and 2010/11, of which 3% is due to reductions in regulated fares in real terms. This is consistent with our London modelling work (see paragraph 27 below). We have not been able to model successfully the major fluctuations in commuting demand since the mid-1980s (see Figure 4).

Overall baseline forecast

14.Taking the commuter and non-commuter demand together, we expect passenger rail demand to continue to grow strongly over the next ten years, by 34%. But this increase assumes that the overall quality of rail travel remains essentially unchanged. It does not take account of any improvements in, for example, journey times, punctuality, reliability and station facilities. As we explain in Chapter 3, investment in quality improvements will induce additional demand, taking total growth under the Plan to 51%.

15.Many lines are already operating at or close to capacity, particularly in the south-east during peak hours and on the West and East Coast Main Lines. In the absence of the investment in the Plan, trains would become more crowded and demand growth would therefore be constrained by a lack of capacity. Our baseline forecast (i.e. without the Plan investments) assumes that passenger growth is constrained to 23%. This reflects our judgement on the proportion of additional demand that could in practice be accommodated[7].

16.Figure 2 below shows the past trends, and the baseline and unconstrained growth forecasts. The challenge that the Plan addresses is to invest in our railways to provide the capacity to meet not only the underlying increase in demand, but also the additional demand that will be generated by improved service quality.

Figure 2: Actual and Baseline forecast total passenger rail demand in Great Britain (passenger kilometres)

Figure 2: Actual and Baseline forecast total passenger rail demand in Great Britain (passenger kilometres)

Modelling rail freight demand

17.Figure 3 below shows that rail freight volumes and market share in Great Britain experienced a sharp decline until the mid-1990s. Since then, volumes have risen by over 40%, and market share has returned to mid-1980s levels. This reflects the joint efforts of the new rail freight operators, Railtrack, and Government (in particular through an increased budget and streamlined procedures for freight grants) to attract more traffic to rail.

Figure 3: Rail freight volumes and share of total freight tonne kilometres (Great Britain)

Figure 3: Rail freight volumes and share of total freight tonne kilometres (Great Britain)

18.Consultants (MDS Transmodal) have carried out modelling work for the sSRA to forecast future demand for rail freight. The size of the overall freight market, both in terms of number of journeys and length of journeys, is determined principally by economic growth: more production and consumption mean more goods being moved across the country. The other key driver of future demand for rail freight is rail's attractiveness relative to other modes, particularly road haulage.

19.Relative cost and service quality are critical in determining whether firms transport their goods by road, rail or another mode. This is reflected in the structure of the model. The costs of rail use can change through:

  • efficiency improvements on the part of freight train operators;
  • investment in rail infrastructure that allows heavier, longer, and higher loads to be carried;
  • changes in track access charges; and
  • where there is a value for money case, Government support through freight grants.

20.Rail freight service quality can change through:

  • improved reliability;
  • increased infrastructure capacity, allowing freight to move more easily to the right places at the right time;
  • better customer focus on the part of freight train operators; and
  • provision of new rail freight terminals.

21.We have modelled the impact of different levels of Government and industry action on these key drivers to improve rail freight's relative competitiveness. We have also taken into account changes in the road freight market and in the costs of road use consistent with the assumptions made in our road traffic modelling (see paragraph 32 below).

22.Our baseline scenario assumed that no major investment in rail freight is undertaken and freight grant provision remains at current levels. With no improvement in rail costs or service quality the model forecasts growth in demand of 38%. Growth is focused in the markets where rail performs best: long distance movements and high tonnage movements. However, this forecast does not take into account capacity constraints resulting from strong forecast growth in passenger rail demand and the move within the logistics industry to larger containers which cannot, without gauge enhancements, be cost-effectively carried on rail routes to ports. Adjusting the model's forecasts to take account of these constraints produced a baseline forecast of 10% growth by 2010, implying a small reduction in market share to 6.5%.

23.The challenge addressed by the Plan is to promote further rail freight growth beyond the baseline forecast as part of a package of measures to secure more efficient and sustainable distribution of goods and services.

London modelling

24.Figure 4 below shows trends in peak travel into central London[8] over the last three decades. Total peak travel has fluctuated in line with the economic cycle. Since the early 1990s there has been a steady increase in both surface rail and Underground travel, but little change in car travel.

Figure 4: Peak travel into central London by mode of transport

Figure 4: Peak travel into central London by mode of transport

25.The total volume of motor vehicle traffic in London as a whole has remained static during the 1990s, but traffic levels in outer London have continued to increase. Average traffic speeds during peak and daytime off-peak periods have been falling gradually. Our analysis using the NRTF modelling framework indicates that congestion levels in London are three and a half times the England average (see paragraph 34 below).

26.The London Transportation Studies (LTS) model[9] has been used to assess the impact of the London element of the Plan on:

  • public transport journey times;
  • Underground crowding; and
  • road traffic volumes and congestion[10].

27.Our baseline forecasts for 2010 assume (% changes are comparisons with 2000):

  • London population and household growth of 4% and 8% respectively;
  • total London and central London employment growth of 4% and 5% respectively towards the middle of the range of independent forecasts;
  • car ownership growth of 12%, mostly in outer London;
  • no increase in public transport fares in real terms; and
  • no significant improvements to public transport services in London[11].

28.The forecasts for key outcomes under baseline assumptions are:

  • Public transport journey times
    Average peak journey times (in terms of time taken to travel one kilometre) increase very slightly, by 0.2%, across public transport as a whole. Average peak journey times on the Underground increase by 2%.
  • Underground crowding
    The number of 'crowded' links would increase from 53 in 2000 to 63 in 2010.
    The number of 'very crowded' links would increase from 20 to 33[12].
  • Road traffic and congestion
    Traffic in central London grows by 2.5%5% between 2000 and 2010. In outer London traffic growth is above 10% in many places. Congestion across London as a whole already three and a half times the national average rises by 13%.

29.The Plan sets out the measures proposed to address these challenges of rising road and public transport congestion.

Road traffic and congestion modelling

30.Total road traffic in Great Britain has grown by over 70% over the last twenty years. This largely reflects rising incomes and increased car ownership leading to increased car use. Goods vehicle traffic has also grown strongly as the economy has grown. The growth in traffic has led to increasing congestion on the road network, particularly in urban areas.

31.Figure 5 below compares road traffic growth and GDP growth. In recent years, however, the relationship between traffic growth and economic growth has eased. It has averaged less than 1:1 since the 1989 peak of the economic cycle. This reflects in particular:

  • higher levels of car ownership and hence reduced scope for further growth amongst those household groups, for example the elderly and young adults, who previously had low levels of car ownership;
  • rising congestion; and
  • increasing application of policy measures that increase the attractiveness of public transport, walking and cycling.

Figure 5: Road traffic and GDP growth

Figure 5: Road traffic and GDP growth

32.We have used the NRTF framework to forecast the impact of policies on traffic growth and congestion in England in different areas and on different road types (and on road traffic emissions see paragraph 39 below). A more detailed description of these models and how they have been developed is at Annex B. In summary, the key stages are:

  • forecasting the underlying growth in car traffic from changes in car ownership and car use by household type;
  • forecasting the underlying growth in goods vehicle and van traffic by commodity sector from projections of trends in load factors, empty running, length of haul, and vehicle type split;
  • changes to these underlying forecasts to take into account responses to increased journey times due to congestion and changes in money costs (for example fuel costs and congestion charges);
  • changes to the forecasts to reflect the impact of other policy measures, including:
    • improvements to the attractiveness of public transport and rail freight (using, for example, the outputs of the LTS, passenger rail and rail freight models);
    • land use planning policies; and
    • sustainable distribution policies, influencing, for example, goods vehicle load factors and empty running.

33.There is currently no widely accepted definition of congestion. The definition used for our analysis of the Plan is the same as for our report Tackling Congestion and Pollution [13]. It is based on the difference between the travel times road users would achieve in free-flowing traffic conditions[14] and the travel times they are forecast to encounter at the levels of traffic, road capacity and hence speeds forecast in the model. This measure is expressed in terms of the average time lost per kilometre driven. We will be giving further thought to measures of congestion as part of our discussions with local authorities, business and others on congestion benchmarks and targets promised in Tackling Congestion and Pollution.

34.Figure 6 compares our estimates of the current levels of congestion in different areas and on different types of road[15]. This illustrates that the greatest problems of congestion are currently in our largest cities. This variation needs to be borne in mind when considering the forecasts of congestion growth for different areas and roads presented later in this paper.

Figure 6: Estimated congestion in 2000 by area/road type compared with all-England average

 

All Roads

Inter-urban
Trunk Roads

 

All areas

London

Conurbations
and Large
Urban

Other Urban

Other

 

Index of time lost per km
(All road average = 100)

100

367

212

98

35

57

35.We prepared baseline forecasts of 2010 traffic and congestion[16] before taking into account the effect of the measures in the Plan. These are set out in Figure 7 below.

36.The underlying assumptions for these baseline forecasts are set out in detail in Annex B. The key points are:

  • no change in car ownership costs, non-fuel running costs or fuel duty in real terms;
  • improvements in new car fuel efficiency sufficient to deliver the EC voluntary agreements with car manufacturers[17]. Together with underlying fuel efficiency improvements in the car fleet and the other assumptions, this results in an average reduction in motoring costs per car kilometre across the car fleet as a whole of some 20% in real terms between 2000 and 2010;
  • no significant change in the quality of public transport;
  • completion of the 37 trunk road schemes in the Targeted Programme of Improvements as set out in the Government's roads review document[18] (but not the three schemes subsequently added to the programme).

Figure 7: Baseline 2010 traffic and congestion forecasts (% changes on 2000, England)

 

All Roads

Inter-urban
Trunk Roads

All areas

London

Conurbations
and Large
Urban

Other Urban

Other

Traffic

22%

14%

16%

21%

24%

29%

Congestion

15%

13%

15%

15%

36%

28%

37.The key points in the baseline forecasts are:

  • traffic growth is lower in urban areas, in part because congestion is already very high;
  • traffic and congestion rise most rapidly outside urban areas, with congestion rising by 28% on the inter-urban trunk road network; and
  • overall traffic growth (22%) is lower than forecast GDP growth (26%), which implies a ratio of around 0.8:1.

38.The Plan addresses the challenge of tackling high existing levels of congestion, particularly in our largest cities, and forecast growth, particularly on the inter-urban network.

Road and rail emissions modelling

39.The NRTF modelling framework now includes an emissions module to forecast the emissions of key pollutants from road vehicles. A new rail emissions model has been developed, described in Annex C. These models have been used to forecast the net impact of policy packages on surface transport emissions in England.

40.In addition, we have used these forecasts to produce estimates of the impact of the Plan on concentrations of local air pollutants taking into account emissions from non-transport sources in the UK and elsewhere. The national scale empirical model we used is a development of that described in the Air Quality Strategy and supporting technical documents. It has been updated to include the most recent air quality modelling results.

Baseline CO2 forecasts

41.Figure 8 below shows the forecast changes in road and rail emissions in England between 2000 and 2010 under baseline assumptions.

Figure 8: Forecast road and rail CO2 emissions in England (2000 and 2010 baseline)

MtC

Road Traffic
CO2 Emissions

Rail
CO2 Emissions

Total

2000

30.3

0.7

31.0

2010 BASELINE

31.0

0.7

31.7

42.Improvements in fuel efficiency, especially in the car fleet where full delivery of the EC voluntary agreements has been assumed, result in overall road traffic CO2 emissions rising by only 2% (0.7 MtC) despite a 22% increase in traffic. The effect of growth in rail use is also offset by increased train loadings and the introduction of more fuel-efficient trains. Total road and rail emissions rise by 2%[19]. The challenge addressed by the Plan is to deliver reductions on this baseline forecast to contribute to meeting our climate change targets.

Baseline air pollution emissions

43.Figure 9 below shows the forecast changes in road and rail emissions in England of two important air pollutants dealt with in the Air Quality Strategy and in recent EC Directives: oxides of nitrogen (NOx) and particulate matter (PM10).

Figure 9: Forecast road and rail NOx and PM10 emissions in England (2000 and 2010 baseline)

 

NOx Emissions

PM10 Emissions

 

Kt

Road Traffic

Rail

Total

% Change on 2000

Road Traffic

Rail

Total

% Change on 2000

2000

487

14.4

501

-

19.7

0.63

20.3

-

2010 BASELINE

198

15.4

213

-57.5%

10.5

0.62

11.1

-45.3%

44. These forecast reductions are dominated by the expected impact of tighter road vehicle emission standards. Road traffic emissions are estimated to fall by 59% (NOx) and 47% (PM10). The challenge addressed by the Plan is top do more in those urban areas where air quality objectives will not be met without further action.

Chapter 3:
Analysis of policy options

45.This chapter sets out our analysis of the expenditure and measures that are needed to address the challenges highlighted in Chapter 2. They are addressed through an integrated package of measures. We have examined the relative cost-effectiveness of six key components of the Plan in tackling two of those challenges road congestion and CO2 emissions. The results of this cost-effectiveness analysis are presented at paragraph 76 below. This is clearly only a partial assessment of the contribution of particular measures to the overall challenges we face. It was nevertheless helpful in informing the overall composition of the Plan and identifying the key choices that may need to be made in the future.

46.As well as presenting an overall assessment of the impact of the Plan, we include an assessment of the impact of three illustrative scenarios, individually and together. The background to these scenarios is explained in Chapter 9 of Transport 2010: The 10 Year Plan, and further details of our assumptions are in paragraph 64 below. They do not represent Government policy or local authorities' views. They illustrate the potential impact of some future changes and policy choices:

  • constant motoring costs in real terms (compared with falling costs under Plan assumptions) and additional transport investment;
  • wider take-up of local charging powers; and
  • introduction of limited inter-urban charging.

Passenger rail

47.The challenge that the Plan addresses is to invest in our railways to provide the capacity to meet not only the underlying increase in demand, but also the additional demand that will be generated by improved service quality.

Demand forecasts

48.Our Plan forecast is that total passenger rail demand will grow by 51% between 2000/01 and 2010/11, around 4.2% a year on average[20]. This forecast includes 36% growth expected as a result of economic growth, reductions in fares in real terms and car traffic growth[21]. Substantial improvements in service quality resulting from the investment in the Plan are expected to generate a further 15 percentage points of demand growth:

  • Journey time improvements. Based on established industry assumptions, improvements in journey times (taking into account waiting time) are expected to have almost a one for one impact on demand for the services concerned, increasing total demand by 8 points; and
  • Improvements in service punctuality and reliability, improved station facilities and other programmes of smaller enhancements are also expected to increase demand, although it is difficult to quantify this effect. We have assumed that this contributes a further 7 points.

49.Using a methodology similar to that used for the Tackling Congestion and Pollution report, we estimate that just under two-thirds of the increased passenger rail demand due to fares reductions and service quality improvements will be from car users switching to rail[22]. We have taken this into account, after adjustment to an England-only basis, in our road traffic forecasts (see Annex B, paragraph 7).

50.Figure 10 below provides a summary of the growth we expect the Plan to deliver, compared with the baseline case, showing the key assumptions in each scenario on car traffic growth and on regulated and unregulated fares. The forecasts of total traffic growth are set out in detail in paragraph 63 below. It is important to emphasise that the Plan road traffic forecasts reflect the impact of the whole package of measures in the Plan and not just improvements in passenger rail service quality.

51.We also illustrate a scenario in which passenger rail demand grows by 83%. This forms part of the illustrative constant motoring costs scenario described in more detail in paragraph 64 below. The higher cost of motoring under this scenario, compared with the Plan forecast, reduces forecast car traffic growth and hence, using our rail demand model, increases passenger rail demand. But higher growth still could potentially be achieved through further fare reductions in real terms in price-sensitive markets, where new rail travel would be most likely to result in significant switching from car. This includes inter-urban markets and also London and south-east off-peak travel. For this illustrative scenario we assumed that further fare reductions start in 2005/06, reflecting the need for additional infrastructure capacity to be in place first. We estimate that inter-urban and London and south-east off-peak fares would have to fall by 4% per annum in real terms thereafter in order to achieve total growth of 83%.

52.The additional public support that would be required for this option is likely to be substantial. It comprises the net cost to train operating companies of lower fares in real terms, and support for the costs of further expansion of infrastructure capacity. We have assumed that these additional support costs would be met by Government.

Figure 10: Passenger rail demand growth forecasts 2000/1 2010/11 (passenger km, Great Britain) and key outputs and assumptions

Scenario

Passenger Rail Demand Growth

Key Outputs and Assumptions

 

Car Traffic Growth in England

Regulated Fares

Inter-Urban Fares - Unregulated

Baseline

23%

21%

RPI-1

RPI

Plan

51%

17%

RPI-1

RPI

Illustrative constant motoring costs scenario (see paragraph 51)

83%

12%

RPI-1/RPI-4

RPI-1/RPI-4

For related documents, pages and internet links, see the column on the right.