Introduction

In this section some case studies to demonstrate the empirical evidence of the use of light rail schemes as policy instruments will be described. Schemes to be examined include Manchester Metrolink and Sheffield Supertram, both in Britain. Contribution to key objectives and alleviation of key problems for Manchester Metrolink, Sheffield Supertram and other systems is summarised below, with more detailed discussion of each following.

Objective	Manchester Metrolink	Sheffield Supertram	Other systems

Contribution to alleviation of key problems
Problem	Manchester Metrolink	Sheffield Supertram	Other systems
Congestion-related delay
Congestion-related unreliability
Community severance
Visual intrusion
Lack of amenity
Global warming
Local air pollution
Noise
Reduction of green space
Damage to environmentally sensitive sites
Poor accessibility for those without a car and those with mobility impairments
Disproportionate disadvantaging of particular social or geographic groups
Number, severity and risk of accidents
Suppression of the potential for economic activity in the area

Manchester Metrolink

Manchester Metrolink Context

More information about the system can be obtained from GMPTE's website (http://www.gmpte.com/travelin/metrolin.htm) and from the LRTA's website (http://www.lrta.org/Manchester/metrolink.html). Maps of the system are available at http://www.gmpte.com/travelin/servicem.htm.

The effects on supply
The total length (39.2km) is small compared with the total road length of 8413 km (DETR, 2001a) in Greater Manchester. The quantities of loaded vehicle-km by public transport in Greater Manchester are shown in loaded light rail km, bus km and train km in Greater Manchester (millions). This shows the quantity of service offered to the public by each mode in a year. It can be seen that the opening of Manchester Metrolink increased the supply of public transport by about 1.5% when it opened. It can be argued that it did not add significantly to the overall supply, but it has probably added significantly in the corridors that it serves.

Source: DETR (2001b) (light rail), GMPTE (2001) (bus and rail).
Note: The figures for light rail are for financial years, which have been allocated to the year in which most of the financial year lies; the figures for train are only for local services supported by GMPTE.

According to calculations by Babalik (2000) Manchester Metrolink uses 28% of its total capacity, calculated as the ratio of the average number of passenger trips per hour to the total passenger carrying capacity of the system per hour. The highest value out of eight systems examined in Britain and North America was 52% for the Tyne and Wear Metro, and the lowest was 13% for Sheffield Supertram.

The effects on demand
The demand for travel by public transport in Greater Manchester is shown in number of journeys by light rail, bus and train in Greater Manchester (millions). It can be seen that total demand for public transport in Greater Manchester has generally declined during the 1990s. Patronage on Metrolink was 8.1 million in its first year of operation, after which it grew to about 12-13 million where it seems to have stabilised. Patronage on other rail services in Greater Manchester has been fairly static. The fact the Metrolink overtook other rail in terms of patronage shows that the latter is not a very important mode in Greater Manchester. Bus is the dominant public transport mode and it is generally declining. Even though it is likely that some users of Metrolink formerly used the bus, Babalik (2000) showed that the introduction of Manchester Metrolink did not seem to alter significantly the long-term downward trend in bus patronage in Greater Manchester. This is partly because bus has such a large share of the market. Even by 1998/99 Metrolink only had 5% of the market compared with 90% on the buses.

Source: DETR (2001a,b)

An alternative way of trying to see the impact of Metrolink on the use of other public transport modes is to compare what happened when it was opened with the trends in comparable areas. Number of journeys in other metropolitan areas outside London, 1991/2 - 1992/3 shows the changes in the numbers of public transport trips between 1991/2 and 1992/3 in other metropolitan areas. Total public transport trips declined by 7% in the other areas, compared with a 1% decline in Manchester, suggesting that Metrolink may have helped to sustain public transport patronage in Manchester. Conversely, train patronage in Manchester went down by 16%, whereas it only went down by 3% elsewhere, suggesting that Metrolink may have attracted some users from heavy rail services. (The heavy rail lines to Bury and Altrincham closed in August and December 1991 respectively, so this partly explains the decline in Manchester). Bus travel in Greater Manchester declined by 3% over this period compared with a 7% decline elsewhere, confirming the point made previously that Metrolink has not had a serious detrimental effect on buses in Greater Manchester.

Source: DETR (2001a)
Note: the other metropolitan areas are West Midlands, Merseyside, South Yorkshire, West Yorkshire and Tyne and Wear.

As well as trips, the total distance travelled can be considered, as shown in number of passenger-km by light rail, bus and train in Greater Manchester (millions). It can be seen that in 1998/99 Metrolink had 12% of the market, heavy rail 15% and bus 76%. The total demand for public transport has declined over the 1990s, with bus declining fast, heavy rail between 210 and 220 million in most years, and Metrolink growing steadily. The faster rate of growth in total distance travelled than the number of trips by Metrolink implies that the average trip length is increasing.

Source: DETR (2001a,b)

It can be seen that the opening of Metrolink coincided with a decline of 2% in total public transport patronage in Greater Manchester. This compares favourably with a 5% decline in other metropolitan areas (see number of passenger-km in other metropolitan areas outside London, 1991/2 - 1992/3). It should be borne in mind that this was a period of economic recession in Britain. Total rail patronage in Greater Manchester grew by 12%, compared with a static position elsewhere, which suggests that Metrolink helped rail travel to grow in Greater Manchester. Bus showed a 9% decline in Greater Manchester compared with a 6% decline elsewhere. Given that the number of bus trips in Greater Manchester went down less than elsewhere, this suggests that a number of longer bus trips have been lost to Metrolink, but there may be some more short trips being made by bus, possibly because of increased seat availability because of the transfer of some longer trips to Metrolink.

Source: DETR (2001a)
Note: the other metropolitan areas are West Midlands (bus only), Merseyside, West Yorkshire and Tyne and Wear.

It is possible to see how much Metrolink contributes to meeting the total travel demand by mechanised modes. As number of passenger-km in Greater Manchester by car, light rail, bus and train, 1998 shows, it is only about 1%. Car is overwhelmingly dominant, with 91% of the market. Public transport has only 9%. Hence, in overall terms Metrolink is making a very minor contribution to meeting travel needs in Greater Manchester. However, by its nature, light rail is very location specific, so it will contribute much more than this in the corridors it serves.

Source: DETR (2001a,b)
Note: the Metrolink figure is actually for the financial year 1998/9. The car figure is based upon the annual road traffic on main roads figure of 11 billion, of which 80% are cars and assuming a car occupancy of 1.54, which is the national average, based on figures from DETR (2001b).

This localised effect of Metrolink on the corridors it serves is illustrated in change in rail demand in Greater Manchester corridors, 1990-93. The changes in rail demand in the Bury and Altrincham corridors are compared with adjacent corridors. The Altrincham corridor shows a 63% increase in the peak and 166% increase in the off-peak. This compares favourably with a 15% decline in the peak and a 3% growth off-peak in adjacent corridors. The Bury corridor is not so buoyant with a 3% decline in the peak and 101% growth off-peak. This can be compared to a 21% decline in adjacent corridors in the peak and a 109% growth off-peak in adjacent corridors.

Source: Table 3.1 in Oscar Faber (1996a)
Note: The Northern and Southern Corridors exclude the Bury and Altrincham corridors.

According to Law et al (1994) patronage was higher on Metrolink than the former heavy rail lines because of:

• Higher service frequency;
  
• Better penetration of the city centre;
  
• The fare structure on Metrolink made many journeys cheaper;

The peak period patronage on Metrolink on the Bury line was lower than anticipated for two reasons:

• Price competition from buses;
  
• Higher fares than charged on the heavy rail.

It is relevant to consider where the patronage on Metrolink has come from. Comparison of estimated observed and forecast sources of Metrolink patronage shows the estimated observed transfer from the monitoring study carried out by Oscar Faber (1996a,b). It can be seen that the majority have transferred from rail, mainly the heavy rail lines that Metrolink replaced. Just over one quarter have come from bus, and about 13% from car. This table does not include any trips generated as a result of the existence of Metrolink. The table also shows the original forecasts of the proportions. A comparison of the two sets of figures suggests that the transfer from car and bus was underestimated in the forecasts and that from rail was overestimated.

Source: Table 5.3 in Oscar Faber (1996a)

An alternative calculation of the modal origins of the Metrolink trips from the University of Salford Monitoring Study is shown in estimated annual Metrolink patronage (millions) by previous mode. This makes the comparison with the situation that was expected to have occurred if the Bury and Altrincham lines had still been operated as heavy rail. This is used rather than the 'before' situation because there was a gap of several months when neither heavy nor light rail operated on these lines and a high quality bus service was operated, which may have influenced travellers' modal choice in the medium term. They estimate that there are 4.5 million more trips on Metrolink than would have used the heavy rail lines that they replaced. Of these, 2.6 million (58%) were previously car trips, 36% were bus trips, 4% used other rail lines, and 4% were not made previously.

Source: Table 2 in Knowles (1996) from the Metrolink Impact Rail User Survey 1993.

Whilst there seems to have been quite a large transfer to Metrolink from the car, this does not necessarily mean that there will be a significant decrease in traffic flows because some people who were previously deterred from using their cars because of congestion may start using them. According to Law et al (1994) there is evidence that car traffic has reduced in the Bury and Altrincham corridors, except in the peak period in the Altrincham corridor, where there has been little change. The effects are complex, but at that time (1993) it seemed reasonable to conclude that there had been some reduction in car use on roads parallel to Metrolink, but it was impossible to measure the effect precisely.

Oscar Faber (1996a) looked at the effects on highway demand in the city centre, as shown in city centre impacts of Metrolink on highway demands. They concluded that there had been a 1.8% reduction in the number of cars entering the city centre in the morning peak and a 0.7% decrease off-peak. They also concluded that there has been a reduction in the number of parking acts: 690 long-stay and 520 short-stay.

Source: Table 6.5 in Oscar Faber (1996a)

More recently Scheurer et al (2001) claim that Metrolink has taken 2.5 million car trips a year off the roads, equivalent to a 10% reduction in traffic on the Metrolink corridor (but possibly releasing space for other car drivers, so that there might be no visible effect on traffic levels). According to GMPTE (1995) Metrolink may have affected the pattern of car purchases in the area it served because between 1991 and 1994, the number of cars per person dropped by 3% in the Metrolink corridor compared with a rise of 5% in the county as a whole.

It is clear that Manchester Metrolink appears not to have had much impact on Manchester. However, it is in operation, it carries over 14 million travellers each year, and it has attracted some motorists out of their cars. It should be borne in mind that it opened at a time of economic recession, so that not much development would have been occurring then. It should also be noted that the two monitoring studies were carried out in the few months after it opened, and much may have happened since then.

Manchester is hosting the Commonwealth Games in 2004, and a modern public transport system is essential in making a successful bid for this type of activity, which in turn can lead to huge amounts of money coming into the city and various developments. Putting it another way, this is a good illustration of the catalytic effect that a light rail system can have alongside many other elements. It is very difficult to unravel the contribution of the individual elements, but they all need to be there.

According to Law et al (1994) the following features have reduced the potential impact of Metrolink:

• Metrolink and bus services were not integrated (because of the deregulated regime for bus services);
  
• No traffic management restraint initiatives on roads parallel to Metrolink;
  
• Car parks not provided at all Metrolink stations;
  
• Since 1992 car parking charges have been levied near Metrolink stations;
  
• Plenty of car parking available in the city centre;
  
• Land rezoning near stations to ensure land use aims were met, was not carried out;
  
• Land was not bought near stations with the aim of having it redeveloped;
  
• No specific grants offered to encourage development.

Sheffield Supertram

Sheffield Supertram Context

More information about Sheffield Supertram can be obtained from the website http://www.supertram.com/info.html. Information can also be obtained from the LRTA website at http://www.lrta.org/sheffield.html. Maps of the system can be obtained from South Yorkshire Passenger Transport Executive's website at http://www.sypte.co.uk/more/maps/supertram.html.

The effects on supply
Opening a new light rail system adds directly to the supply of public transport in an area. The total length of 22 km is comparable with the 146 km of heavy rail route supported by SYPTE (SYPTE, 2000), but it is small compared with the total road length in South Yorkshire of 5851 km (DETR, 2001a).
The effects on demand

The demand for travel by public transport in South Yorkshire is shown in number of journeys by light rail, bus and train in South Yorkshire. Patronage on Supertram has not been as high as expected (Fox, 1996), but it has increased steadily as remedial action has been taken. Total public transport demand in South Yorkshire declined throughout the period shown, and the opening of Supertram has not reversed this trend, but it might have slowed it down, since the decrease levelled off in 1993/94 to 1995/96. Bus patronage has been in long-term decline, and it is not obvious that Supertram has accelerated this trend, a point confirmed by analysis over a longer period by Babalik (2000). Heavy rail demand in South Yorkshire is low, and appears not to have been affected by the opening of Supertram, which is not surprising given the route pattern of Supertram.

Source: DETR (2001a,b)
Note: Rail services are those supported under Section 20 of the 1968 Transport Act.

This can be compared with the shift changes in patronage on bus and heavy rail in other metropolitan areas at the time Supertram was opened as shown in number of journeys in other metropolitan areas outside London, 1993/4 - 1994/5. In the other areas there was a small growth in bus use whereas in South Yorkshire there was a small decline, suggesting that Supertram may have prevented a short-term growth in bus patronage in South Yorkshire which was probably associated with the improving economic situation at the time. Heavy rail showed a decline in the other areas whereas it was about constant in South Yorkshire at a very low level.

Source: Department of the Environment, Transport and the Regions (2001a)
Note: the other metropolitan areas are West Midlands, Merseyside, West Yorkshire and Tyne and Wear (Greater Manchester has been excluded because of the introduction of Manchester Metrolink).

Abstraction of Supertram trips from other modes shows where Supertram trips have come from. It can be seen that most trips (55%) have transferred from bus. 20% have come from car and 12% are new trips that would not have otherwise been made. Given that patronage on Supertram is low, 20% transfer from car would not make a huge difference even if no other travellers started using their cars because of the resulting reduction in congestion.

Source: W S Atkins (2000)

According to calculations by Babalik (2000) Sheffield Supertram uses 13% of its total capacity, calculated as the ratio of average passenger trips per hour to the total passenger carrying capacity of the system per hour. This is the lowest value out of eight systems examined in Britain and North America where the highest was 52% for the Tyne and Wear Metro.

Objective	Comment
	Increased range of transport options in Sheffield, so may be meeting some travel needs more efficiently. Little impact in terms of patronage, so little impact in terms of reducing road traffic.
	No operating subsidy received. Did not significantly alter long-term downward trend in bus patronage in Sheffield (Babalik, 2000). Hence, Supertram has added to the efficiency of public transport in Sheffield in economic terms.
	Small local effect on residential property prices (Crocker et al, undated). Prices fell between 1988 and 1993 in anticipation of construction. After opening, prices rose for property near the system, but not back to the 1988 level. N.b. analysis undertaken just four months after opening. No discrete impact on commercial or industrial property prices detected. No noticeable effect on planning applications and land use change detected.
	Reduced car trips compared to what would have happened will have made Sheffield more pleasant, but effect will have been small. Effects also localised along corridors served.
	Negative impact on Sheffield city centre (Babalik, 2000) – first line opened took shoppers away from the centre to the Meadowhall shopping mall. 35% decline in city centre retail turnover (Rowley, 1995) – trend exacerbated by opening of Supertram. However, fact that the two centres are linked may have boosted patronage.
	Environmental effects small. Concern over visual intrusion of overhead wires at development stage, but addressed in public consultation. Opportunity to improve local streetscape taken during construction.
	Gender make-up of Supertram patrons mirrors that of public transport users as a whole (W S Atkins et al, 2000). Supertram 41% male, 59% female, bus 38% and 62% respectively, all modes 49% and 51% (W S Atkins, 2000).
	Age profile similar to all public transport users. Supertram popular with elderly people provided they can reach a stop because they pay the same concessionary fare as on the bus, but benefit from additional comfort of the vehicle and ease of access/exit (low floor) (W S Atkins, 2000).
	Lowest socio-economic group had trip rate of 2.5 per week, whilst other three groups had a rate of 2.7 per week. One third of Supertram trips by individuals in two highest socio-economic groups, whereas only 25% of bus trips made by members of these groups.
	Vehicles seen as easier to board than buses and having more space to stow pushchairs and bulky items.
	Any reduction in net road traffic as a result of the introduction of Supertram should have improved safety, but effect will have been small and potential localised.
	Supertram runs through area where comprehensive regeneration project implemented by Sheffield Development Corporation. However, poor co-ordination between two schemes. Supertram runs along margin of new development with poor access from development to stops (Lawless, 1999).
	Supertram has had positive impact on city’s image, especially in eyes of external agencies. Also useful in city’s tourist promotion programmes.
	Considerable disruption in city centre and along Supertram route during construction had inverse effect on efficiency and productivity of companies in the locality. New road construction had a stronger impact on industrial and commercial development proposals than Supertram (Crocker at al, undated). 12-15% of land use change in three areas attributed to Supertram, but most development likely to have gone ahead regardless, but may have been brought forward in time. More positive image of Supertram since opening should lead to positive impacts on businesses (Crocker et al, undated).
	Little impact on labour market (Crocker et al, undated). Some evidence on improved access to areas such as Mosborough, served by system, and of people able to job search over a wider area, but effects small. Line 1 might lead to 295 jobs, line 2 between 380 and 1275 (jobs in local economy, not system construction). Application for funding suggested Line 1 would create 1135 jobs and Line 2 3000 (Crocker et al, undated).
	N.b. Supertram opened at a time of economic recession, therefore little movement in local economy or property market at the time.

The overall impacts of Supertram

Because patronage on Sheffield Supertram is low (10.9 million in 1999/2000) it is not likely to have had many impacts. The forecast patronage turned out to be very optimistic which has raised some questions about why this was the case. According to W S Atkins (2000) the two biggest sources of error were assuming that there would be a major transfer from bus to Supertram and assumptions about trips from new developments. The former problem arose from the fact that the rival bus companies decided to operate in a very competitive way, in terms of both routes and fares, which meant that there was not the scale of transfer anticipated. These problems have largely been solved by the taking over of operation of Supertram by Stagecoach which is a major bus operator (It was not the major incumbent operator in Sheffield at the time of the opening of Supertram. Fox (1996) argues that the incumbent operator wished to take over Supertram when it was privatised and so had an interest in it being in a financially weak position). The forecast errors arising from assumptions about new developments were symptomatic of the problem of poor co-ordination between the city planners in Sheffield and SYPTE who were developing Supertram (Fox, 1996). The route pattern was devised to serve some high density developments. The three tower blocks at Herdings Park, which are at the end of a short branch line, were emptied of residents because they were in a very poor state of repair but the line was still built. The Kelvin development which would also provide customers was demolished rather than being renovated as originally planned, and the Norfolk Park Estate has been gradually emptied so that it can be redeveloped at a much lower density.

The annual level of revenue from patronage forecast for 1996 was 22.1 million: 17.1 million on Line 1 and 5.0 on line 2. The actual figure was about 6.6 million at this time (Haywood, 1999). Haywood (1999) gives the following reasons for the shortfall in patronage on Sheffield Supertram:

• Decline in bus use 24%
  
• Competitive buses 12%
  
• Supertram frequencies 8%
  
• Supertram run times 8%
  
• Supertram fares 3%
  
• Park and ride 4%
  
• New developments 4%
  
• Unexplained 4%
  
• Actual patronage 30%

Whatever the reasons for the errors in forecasting patronage on Supertram, it was not a very satisfactory procedure. What is also clear is that patronage is increasing steadily following changes to the service pattern and fares, the introduction of conductors to help overcome problems of vandalism, and improvements in the local economy. Technically, it is a very good system, but the many problems have led to delays in it reaching its potential.

Other systems

Context
Reference has already been made to a number of light rail systems around the World. In most cases, specific monitoring studies have not been carried out, unlike the Greater Manchester and Sheffield systems, so it is not possible to draw detailed conclusions about their impacts. It is, however, possible to take information from the surveys of light rail and similar systems by Mackett and Edwards (1998) and Babalik (2000). The Manchester and Sheffield systems were included in both surveys and so will be included here where appropriate for comparison.

Effects on supply
No detailed information of the effects of the new light rail system on the total supply of transport are available other than those already shown for Greater Manchester and Sheffield. It is likely that in all cases the total did increase, because, only if road space were decreased significantly to allow on-street running, would it be possible for the development of a new system to lead directly to a decrease in transport supply.

Effects on demand
All the systems examined in the two surveys mentioned above are carrying large numbers of passengers, and so have stimulated some demand. One useful indicator of demand is how well actual patronage matches that forecast since the forecast would have been used as part of the planning process and to help determine whether the project would be worthwhile financially. Forecast and actual patronage on a weekday for light rail systems in thousands shows the forecast and actual patronage for a number of modern light rail systems.

It can be seen that there are huge errors in the forecasting procedures. Out of the ten systems shown, patronage was overestimated in four and underestimated in six, with errors of up to 161%. The one Canadian example, in Vancouver, was an underestimate by 36%. On the Manchester Metrolink demand was underestimated by 25%, but as Knowles (1996) showed, the type of patronage forecast was very different to the actual, with much more off-peak travel and much less peak travel in reality than expected. The forecasts for Sheffield Supertram were significantly out (see "The overall effects of Supertram"). Forecasts for the Tyne and Wear Metro were fairly close to the actual values, but the patronage declined after this point, and was down to 126 900 by 1996.

Source: Mackett and Edwards (1998) and Babalik (2000), using information from Pickrell (1990), Dunphy (1995), Warren (1995), Federal Transit Administration (2000) and DETR (2000b).

The four US systems in which patronage was underestimated, in Buffalo, Pittsburgh, Portland and Sacramento, were all constructed using some Federal funding, giving some credence to the claim that patronage demand was often overestimated under these circumstances. On the two other US systems patronage was underestimated: San Diego Trolley which was initially built with no Federal funding and St Louis MetroLink which was constructed after the funding rules were changed.

Babalik (2000) has calculated the extent to which the total capacity of light rail systems is used, as shown in percentage of total capacity used on light rail systems.

The figures look low, in general, because they are averages over the whole day, including reverse flows during peak periods. The highest value is for the Tyne and Wear Metro, but this is high partly because during the construction of the extensions, stations were designed to accommodate only two-car trains instead of the original four in order to save money. If the capacity of the original system were considered then 38% of the capacity would be used. This suggests that the San Diego and St Louis systems are the most efficient in terms of matching supply to demand. They were both systems where the actual demand has exceeded the forecast which probably explains the relatively high capacity utilisation: they have more passengers than they were originally expected to carry.

Source: Babalik (2000)
Note: the capacity used is the ratio of the average number of passenger trips per hour to the total passenger carrying capacity of the systems per hour.

Thus, it can be seen that the systems do all meet the passenger demand to some extent at least. It can also be seen that the demand forecasting procedures have not been very good.

The overall impacts of other light rail systems

There is little doubt that the systems have all added to the supply of transport in their areas. There is limited evidence on their impact on demand. Certainly they are all carrying passengers, in significant numbers in some cases. It is not clear how far they have increased the demand for public transport. They do not seem to have had much impact on traffic levels although there is some evidence that they are attracting some motorists out of their cars. One thing that is clear is that the techniques for forecasting demand have not been very accurate. In some cases this may have been the result of deliberate action by the planners to obtain funding.

Of the various objectives identified 'Contributing to economic growth' is the only one where the light rail systems have had a significant impact. This partly reflects the objectives of developing the systems, because this is one of the most important reasons for developing such systems. Whilst not all the systems have had such impacts and those that have done so, have required the use of complementary policies, this does seem to be the major area in which light rail systems have great potential.