Policy Instruments

High Occupancy Vehicle (HOV) Lanes
SummaryFirst principles assesmentEvidence on performancePolicy contributionComplementary instrumentsReferences

Evidence on performance

Experience in Europe: Leeds, UK

In 1998, UK’s first High Occupancy Vehicle (HOV) lane was introduced on the A647 Stanningley Road and Stanningley By-Pass which form the principal radial route to the west of Leeds city centre and are part of the route linking Leeds and Bradford (see figures below). The scheme was experimental at first but has become permanent. The road experienced severe congestion and there were few public transport priority measures.
The £450 HOV lane scheme covers a total of 1.5km of 2.0km long dual carriageway in two sections. They operate in the morning (07:00 – 10:00) and evening (16:00 – 19:00) peak periods on Mondays to Fridays. Only buses, coaches, other vehicles carrying 2 or more people, motorcycles and pedal cycles are allowed on these lanes (HGVs over 7.5T are not permitted).

Location of the A647 HOV Scheme in Leeds

Scheme layout, High Occupancy Lane, Leeds. Courtesy of Leeds CC.

Illustrations of the Leeds HOV lane scheme.

Courtesy of JMP Consultants Ltd

source ICARO (1999)

Courtesy of Leeds City Council

This project was part of an EU research project called ICARO (Increasing CAR Occupancy). Its objectives were to increase car occupancy by encouraging car sharing; and to demonstrate the feasibility of providing a lane for shared use by buses, other high occupancy vehicles, motorcycles and cycles. More detailed information can be found in the ICARO project deliveries and final report (available at

The following summary information is taken from the DfT’s ‘Bus Priority: The Way Ahead’ initiative (case study on HOV Lanes) (DfT, 2004); Leeds City Council’s HOV Fact Sheet (2002); and ICARO (1999).
Impacts on demand
Prior to implementation of HOV lane, 30% of cars on the A647 (Stanningley Road) had 2 or more occupants. With the inclusion of buses, one-third of all vehicles carried two-thirds of all people (2225 of 3645) in the morning peak period. The journey in free flow conditions could take about 3 minutes, regularly had taken over 10 minutes. Therefore a priority lane such as an HOV lane would benefit a majority of the travellers in terms of journey times. However single occupant drivers (total of 1420) would be expected to suffer some additional delay due to capacity reduction caused by the HOV lane.
The results below based on the ‘Before’ surveys that took place in May and June 1997 and ‘After’ surveys that took place in May and June 1999. Data collected included
• traffic counts in the morning and evening peak periods;
• vehicle occupancy;
• journey times;
• queue lengths; and
• personal injury accidents.
In addition to this, public attitudes and driver behaviour information were analysed from household and roadside interview surveys. Air quality was monitored by an environmental monitoring station on the route.
The following table summarises the results of before and after analysis done by Leeds City Council (LCC, 2002). It was reported that, after an initial reduction, traffic levels gradually increased to its previous levels with about 5% increase in HOVs. This might indicate that there was an exchange of HOV and non-HOV traffic between the A647 and parallel routes. On the other hand, 26% of HOV interviewees were apparently new car pools and cited the HOV lane as the reason for forming them. Relatively low support amongst HOV drivers (about 66%) might have resulted from the fact that these drivers also made peak period journeys as non-HOV drivers. When doing so, they did not benefit from the journey time savings observed.

Leeds HOV lane impacts (Bus Priority Initiative case study, (LCC, 2002))



Morning peak traffic flows

(07:00 - 10:00)

         Immediately after opening 20 % traffic reduction (due to driver avoidance).

         By late 1999, traffic flows returned to prior levels

         Sight increase in scheduled bus services, motorcyclists and cyclists

Evening peak traffic flows

(16:00 to 19:00)

         10 % reduction at scheme inception

         by June 1999, traffic flows returned to the 'before' level

         By June 2002 traffic flow increased by a further 14 %

Occupancy and mode share:

         between 1997 and 1999, HOVs in morning period increased by 5 %

         Average car occupancy rose gradually from 1.35 in May 1997 to 1.43 by June 1999 and 1.51 in 2002

         Bus patronage increased by 1% in the first year of operation (There are indications of further growth in bus patronage since 1998 but no real data available to analyse)

Journey times:

         Morning peak journey time savings for buses and other HOVs were 4 minutes (comparing 1997 to 1999 data)

         Reduction of 1 minutes in non-HOV journey times in the same period.

Queue Lengths

         By giving priority to HOVs, two queues of equal length have been transformed into a long queue in the non-HOV lane and a short queue in the HOV lane

         No evidence of non-HOV queues extending


         Reduction of 30 % in casualties in a period of three years after scheme implementation.


         Lane violation levels were low in the months following implementation

         In 2002, lane violation levels were still less than 6 % despite a relaxation of enforcement.

Public attitudes:

         An increase from 55% to only 66 % in HOV drivers support for HOV lane (results from roadside interviews in 1999).

Air quality:

         Little change in air quality

         A noticeable noise reduction coinciding with both the morning and evening periods of HOV lane operation

The figure below shows the HOV lane journey time change along different sections of A647 that resulted from HOV Lanes and signal improvements. In 1998, morning peak HOV journey time savings were 3½ minutes for a 5km trip from the Leeds Outer Ring Road to the Inner Ring Road. In 1999, the time saving increased to just over 4 minutes. The figure also indicates that the journey time savings starts after the first 2.5 -3 km and changes sharply from time lost to time gain.

A647 AM peak HOV lane journey times

The figure below shows the journey time change on the general purpose lanes (outside lanes) within same sections of A647. In 1999, overall inbound non-HOV journey times did not increase and were total of 1½ minutes shorter in the morning peak for the same 5 km long journey.
A647 AM peak non-HOV lane journey times

Scheme implementation cost was £585,000 at 1998 prices. Following the success of the A647 scheme, Leeds City Council is now planning to introduce HOV lanes on the proposed East Leeds Link Road.

Experience in Europe: Bristol, UK

On the congested dual two-lane Avon Ring Road near Bristol, the local (South Gloucestershire) Council had wanted to implement a bus lane but frequencies were too low to justify reallocating road space to buses alone. It therefore opened an HOV lane for buses, taxis and cars with 2+ occupants. The lane has been extended from the original 750m to 1.2km in length (comprising two sections separated by a roundabout) and operates in one direction in the morning peak only.

The lane has led to an increase in efficiency; the proportion of single occupancy vehicles has fallen from 80% to 68%, and traffic levels have increased by 10% (as a result of vehicles re-routeing from parallel roads) as the lane has ‘smoothed’ flows and allowed higher throughput.

Journey times for all vehicles have fallen from 20 minutes to 6 in the HOV lane and 12 in the mixed use lane.

Experience in Europe: Madrid

One of the first and more significant European HOV lanes is the 16 km two-lane scheme on the median of the N-VI motorway into Madrid which was opened in January 1995. It was introduced when the motorway was widened to accommodate traffic from nearly affluent suburbs. The lanes are reversed to match peak flows and this has greatly increased throughput of traffic.

A study of operational performance undertaken by the Madrid Polytechnic University ( demonstrated that:

In terms of efficiency, the N-VI HOV lanes were found to be positive as they carry 59.3% of the morning peak hour travellers using 2 lanes, while the 3 lanes of the main roadway carry only 40.7% of the travellers.

The proportion of 2+ vehicles increased from 30% in 1991 to 40% in 1996
The lanes have attracted a growth in public transport mode share, rather than ridesharing.

The HOV lanes retain high levels of service through most of the peak period because they still have some spare capacity. To fully utilise all available road space, active policies to promote carpooling and public transport use are needed.

Public transport use increased by 40% in the period 0700-1000 in the year following implementation and more bus operators have begun operating, so frequencies are higher. Monitoring reports also note a fall in congestion, but there are no actual measures given.


Other results included:
• In inbound Madrid, the average occupancy in cars has increased from 1.36 passengers per car in 1991 to 1.67 passengers in 1997.
• The modal split in the corridor has changed so that the use of buses has increased from about 17 % in 1991 to about 26 % in 1997. In the same period the use of cars has decreased from about 56 % to about 48 %.
• Reduced travel times in and out of Madrid has lead to the change in modal shift.

Experience in Europe: Trondheim
The Trondheim scheme, which opened in May 2001, uses one lane in each direction on a dual two-lane radial on the approach to the city centre. The route was characterised by congestion during the morning and evening peaks which affected the reliability of public transport operations, but the delays were light – typically 5 minutes per vehicle lasting for a period of about half an hour.
Different Trondheim schemes were tested with the American CORSIM model and a 2+ scheme selected to balance usage and speed in the HOV lane and congestion to general traffic.

Nordic Road and Transport Research (No 3, 2001) reports that the HOV lanes offer journey time advantages of only 2-3 minutes in the peaks and yet the proportion of 2+ vehicles has increased from 30 to 40% and there has been a corresponding decline in single occupancy vehicles from 70 to 66%. This survey was limited to the HOV corridor and it was not clear if carpoolers have been attracted from alternative routes.

Traffic levels in the area have remained stable, possibly because some traffic has been attracted from surrounding residential streets, while the number of people using the route has increased and person hours in the morning peak have fallen by 20%.

In the mixed use traffic lane, the congested period has increased from half to three-quarters of an hour; however, the city authority considers this to be acceptable, especially has the scheme helped to regulate stop-start conditions downstream.

Other European examples
As well as Leeds, Bristol, Madrid and Trondheim, there are converted bus lanes in Linz (Austria) and rural Stockholm. There were also schemes in Amsterdam and Berlin that were withdrawn because of legal opposition,

Shirley Highway and Santa Monica Freeway, USA

The first major US schemes were the Shirley Highway near Washington DC, which opened in 1973 with two HOV lanes for cars with four or more occupants, and the conversion of one lane of the Santa Monica freeway in Los Angeles to HOV 3+ operation (including buses) during the peak hours.
The successful Shirley scheme now carries half of the commuters in the corridor at an average speed of twice the mixed flow lane. However, the Los Angeles scheme reverted to mixed use after just four months because of poor planning and local opposition.

Adverse side effects
When it is not combined with complementary measures such as travel plans, car sharing initiatives, public transport improvements, it might result in under-utilisation in road capacity and therefore cause extra delays to general purpose lane users.
In contrast, when time saving are high, it can induce trips from parallel routes, induce additional trips or shifts from other times.
It may also cause abstraction of public transport users to join SOV drivers to enable them to use the HOV lane.

Top of the page

Text edited at the Institute for Transport Studies, University of Leeds, Leeds LS2 9JT