Accidents at Signal Controlled Junctions in Glasgow - Research Findings

DescriptionThis report was to investigate whether the use of red light cameras in Glasgow District has created wider benefits, including a reduction in the number and type of accidents.
ISBN0 7480 5504 5
Official Print Publication Date
Website Publication DateDecember 29, 1998
Development Department Research Programme Research Findings No. 23 (1996)
Accidents at Signal Controlled Junctions in Glasgow

Halcrow Fox

ISBN 0-7480-5504-5Publisher The Scottish officePrice £5.00
Recent research has found that the installation of camera equipment at signalised junctions in Glasgow has resulted in a reduction of red light running and associated accidents at the camera sites. The Scottish Office commissioned Halcrow Fox to investigate whether the use of red light cameras in Glasgow District has created wider benefits, including a reduction in the number and type of accidents at all signalised junctions across the whole of the district, and not just at the specific camera site localities. The period studied covered 3 years before and 3 years after the installation of cameras.
Main findings
  • There was a significant decline in accidents at all signalised junctions in Glasgow between the before and after periods studied.
  • Both personal injury and non-injury accidents caused by red light running fell by about a third.
  • Red light cameras appeared to be only one of several factors contributing to the reduction of accidents at signalised junctions. Red light running accidents accounted for 20% of the decline.
  • Changes in accidents due to other causes were more significant, particularly accidents caused by pedestrians "crossing carelessly" which account for 44% of the reduction in personal injury accidents.
  • Injury accidents caused by red light running declined more sharply at junctions away from the camera sites suggesting that factors such as junction improvement, traffic management and increased vigilance were also important.
  • The reduction in red light running accidents cannot be attributed to changes in traffic flows during the study period.
  • Results from a Cost Benefit Analysis suggest that the economic returns from the implementation of red light camera technology in Glasgow District have been positive.
The research
The Strathclyde Police Red Light Initiative was formulated in the late 1980s with the objective of promoting road safety and reducing road accidents. This was in anticipation of the 1991 Road Traffic Act (eventually implemented in Scotland in 1993) which enabled the use of automatic detection devices in the enforcement of road traffic law. As part of this initiative, red light cameras were installed, and by 1995 there were 8 camera sites at signalised junctions and 3 at pelican crossings in Glasgow District. Previous research has studied the link between red light running and accident causation and found that 17% of personal injury accidents at signalised junctions in Glasgow District were primarily caused by red light running. Recent research has found that cameras significantly affected driver behaviour at the camera sites, leading to a large reduction in the incidence of red light running, and a fall in accidents at the camera sites.
To develop this work further, The Scottish Office commissioned Halcrow Fox to investigate the possible wider effects of camera technology on all signalised control junctions in the Glasgow district. Specific objectives of the research included:
  • to determine the characteristics and frequency of accidents at signalised junctions and near pelican crossings for time periods before and after the introduction of cameras
  • to assess the impact of the cameras on the number and type of accidents at all signalised junctions and pelican crossings within Glasgow District
  • to examine the results in the light of national and regional trends in accidents, and of changes in traffic flows in Glasgow District, to determine whether broader trends may be responsible for the observed changes
  • to estimate the benefits and costs of using camera technology in relation to any reduction in the number of accidents.
Methodology
The research was based on a statistical analysis of data which was drawn from different sources, and the study involved three distinct stages; firstly, the collection of accident, traffic and camera related cost data; secondly, the processing and analysis of that data; and thirdly, the use of the results to estimate the costs and benefits of camera technology.
National trend data on personal injury accidents only were drawn from the data series Road Accidents Scotland published annually by The Scottish Office, and were compared to the incidence of personal injury accidents occurring in Glasgow during the study period.
Strathclyde Regional Council provided data on all accidents (injury and non-injury) at junctions and near pedestrian crossings in Glasgow district for the period January 1989 to November 1995, along with data on traffic flow.
Analysis of the accidents at signalised junctions was divided into 3 distinct time periods:

'Before' - January 1989-June 1991, before cameras were introduced.

'Interim' - July 1991-March 1993 when the cameras were in use but before the legislation allowing camera evidence was enacted; warning letters only were issued.

'After' - April 1993-November 1995 when the cameras were used with fixed penalty fines.

Different time periods were used for accidents at pelican crossings, and for comparison purposes the results were presented as accident rates per month.
A cost benefit analysis was made by estimating the cost of purchasing and operating the camera system compared with any reduction in accident costs. This was estimated over a 20 year period and discounted to 1993. Outputs are in terms of 1998 prices and 1993 values.
Results
National trends
Figure 1 below shows that there has been a broad downward trend in personal injury accidents since 1989, but by 1994 the rate of decline in the number of accidents in Glasgow was greater than for Scotland as a whole.
Figure 1
Personal injury accidents and traffic 1989-1995
Figure 1 Personal injury accidents and traffic 1989-1995
Accident trends within the study area are broadly in line with Scottish accident trends, and the overall decline in accidents for Scotland of approximately 20% between 1989 and 1994 compares with a decline in accidents at all signalised junctions in Glasgow of 25% and in accidents at all pelican crossings in Glasgow of 30% over the same period. The substantial drop in the number of accidents in 1993 is not matched at national level and coincides with the introduction into full operation of Strathclyde's first red light camera.
Accidents at signalised junctions in Glasgow district
Over the study period, the number of all accidents at signalised junctions reduced by 2,260, with fatal accidents reducing from 27 in the before period to 8 in the after period, serious injury accidents from 466 to 248 and slight injury accidents from 1,665 to 1,061. The number of non injury accidents reduced from 4,579 to 3,160. Table 1 below describes the changes in accident severity by rates per month. The downward trend in these accidents is reflected in national figures for built up areas.
Table 1
Accident Severity at Signalised Junctions by Time Period (Rate per Month)
Time Period
Accident SeverityBeforeInterimAfterAfter-Before% Change
Fatal
per month

0.8

0.5

0.3

-0.5

-67%

Serious
per month

12.9

8.6

7.8

-5.2

-40%

Slight
per month

46.3

38.5

33.2

-13.1

-28%

Non-Injury
per month

127.2

109.2

98.8

-28.4

-22%

Grand Total
per month

187.1

156.9

139.9

-47.2

-25

Analysis showed that there has been a significant decline of 25% in accidents at all signalised junctions in Glasgow between the before and after periods. In proportionate terms personal injury accidents have experienced a greater decline (-31%) than non-injury accidents (-22%) and the more serious accidents have fallen by the largest percentage.
Personal injury accidents
When personal injury accidents at all signalised junctions are categorised by primary cause, it is found that there has been a reduction of 32% in red light running accidents, but that the greatest decline in personal injury accidents has been in the "crossing carelessly" category, which has reduced by 54% between the before and after time periods. This reduction represents 44% of the total reduction in personal injury accidents overall. Factors contributing to this decline may be the provision and use of safer pedestrian crossing facilities, improved behaviour by pedestrian and driver as a result of road safety education programmes, and changes in traffic and pedestrian activities. The only type of accident which increased over the study period was that caused by "failure to keep distance" which may reflect greater caution on the part of drivers approaching and driving through junctions.
Table 2
Primary causes of PIA's at signalised junctions by time period

Time Period

Primary Cause

Before

Interim

After

After-Before

% Change

Red Light Running
per month

10.5

8.8

7.1

-3.4

-32%

%

17.5

18.0

17.3

Crossing Carelessly
per month

15.3

9.9

7.0

-8.3

-54%

%

25.5

20.7

17.0

Unsafe Right Turn
per month

8.6

7.9

5.1

-2.8

29%

%

14.4

16.5

14.8

Fail to Keep Distance
per month

5.2

7.0

5.7

0.5

+8%

%

10.3

14.7

16.2

Other
per month

19.3

14.3

14.3

-5.1

-26%

%

32.3

29.9

34.5

Grand Total
per month

59.9

47.7

41.2

-18.8

-31%

100.0%

100.0%

100.0%

Note: All primary causes shown for categories where more than 5 accidents per month occurred in the before period.
Table 3 below indicates that in the 'camera present' areas, personal injury accidents caused by red light running fell by 25% at the first 6 sites, and by 40% at the latest 2 sites between the before and after periods. A further reduction of 21% in the adjacent areas may indicate some wider area effect. However, the spatial analysis of the location of accidents in relation to camera sites shows a greater reduction in injury accidents caused by red light running in areas far removed from the camera sites. This demonstrates that other factors such as junction improvements, local traffic management and increased pedestrian and driver vigilance may have been important in reducing red light running accidents across the whole area.
Table 3
Changes in the number of PIAs (per month) at signalised junctions by area of incidence, primary
causation and time period
All PIAsRLR PIAs

Number of
Junctions

AreaBeforeAfter% DiffBeforeAfter% Diff
110.97.9-27.63.22.4-25.453
20.80.4-51.80.30.9-38.93
328.419.8-30.23.93.1-21.1169
420.016.1-19.53.12.1-32.7143
Note: Definition of Areas:
  • Area 1 - all 1km squares which contain one or more of the six camera sites installed in 1990;
  • Area 2 - 1km squares containing one of two camera sites installed in early 1994;
  • Area 3 - 1km squares which are adjacent to areas 1 and 2;
  • Area 4 - the rest of Glasgow District.
In the study of different types of vehicles involved in red light running accidents, the percentage of offences caused by buses and taxis was the only category subject to significant change, with the percentage halving from 16% to 8% after the installation of cameras. This change may reflect a higher level of awareness of the existence and use of red light cameras which would quickly be disseminated amongst groups of professional drivers.
Non-injury accidents
Analysis of non-injury accidents at signalised junctions showed a reduction of 28% between the before and after periods within this group. There has been a reduction of 35% in red light running accidents, but other causes have experienced a sharper decline, with accidents caused by overtaking improperly declining by 67% and those caused by misjudging the distance by 54%.
Traffic flows
The research investigated whether any other 'external' factors other than the introduction of red light cameras were likely to have influenced the trend in accidents since 1989. One obvious variable is traffic flow. Accidents involving more than one vehicle are known to consistently increase with traffic flow, although eventually congestion reduces speed and hence accident severity. Traffic flows in Glasgow District between 1990 and 1995 were analysed, and it was found that traffic volume near to camera sites had increased overall, though with some local decreases. It appears therefore that accident reductions achieved over this period are not attributable to changes in traffic flows; rather they have been achieved in the face of increasing numbers of vehicles on the road.
Cost benefit analysis
Results from analysis of the costs and benefits of the scheme suggest that investment in camera technology within the study area has provided a significant return in terms of reduced accident costs. A number of scenarios were devised which differed by assumed impact of red light cameras on accidents at the immediate camera site and in the wider area. Taking a reasonably conservative view of the scale and scope of benefits, the discount net present value of the project is estimated to have been £1 million in relation to total costs of £452,000. This is based on the following assumptions:
  • inclusion of the 8 signalised junction camera sites
  • inclusion of 50% of savings in red light running accidents at the immediate camera sites
  • inclusion of fatal accident savings and
  • exclusion of any potential wider effects.
Conclusions
Red light cameras are one of several factors contributing to the fall in accidents at signal controlled junctions. Clearly, the existence of cameras at signalised junctions has had an impact on the number of accidents occurring at the immediate locality and findings suggest that there may well be a wider effect. The analysis of Glasgow traffic flow has shown that the reduction in red light running accidents at all signal controlled junctions cannot be solely attributed to changes in traffic flow during this period. The very substantial drop in the number of accidents at all signalised junctions and pelican crossings, which is not matched at the national level, coincides with the introduction and full operation of Strathclyde's first red light cameras in 1993.
This research concludes therefore that red light cameras have a definite beneficial effect on road accidents at the camera sites, are one of several factors effecting a reduction in accidents in the wider area, and provide a positive economic return on their installation and operational costs.
6
Previous Relevant Research
This work completes a programme of evaluation carried out on behalf of the Scottish Office on the link between red light running and road accidents and the impact of red light cameras on accident reduction in the Glasgow area.
"The Strathclyde Police Red Light Initiative: Accident Monitor", Scottish Office Central Research Unit, 1992
This study examined the link between red light running and accident causation before the use of red light cameras. Analysis of the Strathclyde Regional Council accident database for the period 1989 to 1991 showed that 17% of personal injury accidents in Glasgow District were primarily caused by red light running. It was estimated that the introduction of red light cameras at specific sites in Glasgow would be cost effective, taking into account the capital costs required for their installation and maintenance and potential accident cost savings.
"Running the Red", Scottish Office Central Research Unit, 1995
Following the introduction and use of red light cameras in Glasgow District, the effect of red light cameras on driver behaviour and compliance was investigated to determine whether this resulted in a reduction of red light running accidents at the camera sites. Before, interim and after surveys were carried out at six camera and six control sites between 1991 and 1994. It was found that driver behaviour at the camera sites was significantly affected by the presence of red light cameras leading to a decline of 69% in the number of red light running infringements. A separate analysis of accidents at the junction sites revealed a 62% reduction in personal injury accidents over the same period. In addition, there was some evidence that cameras had an influence on junctions in the immediate vicinity of the camera sites.
"Accidents at Signal Controlled Junctions and Pelican Crossings in Glasgow", the Research Report summarised in this Research Findings, is available price £5.00.
Cheques should be made payable to The Stationery Office Books and addressed to:
The Stationery Office Bookshop,
71 Lothian Road,
Edinburgh EH3 9AZ
Telephone: 0131-622 7050 or Fax: 0131-622 7017.

The report can also be ordered online from:www.thestationeryoffice.co.uk

Further copies of this Research Findings can be obtained from:
The Scottish Office Central Research Unit
Area 2J
Victoria Quay
Edinburgh EH6 6QQ
Tel: 0131-244 7562

Page updated: Friday, June 06, 2008