Task Force on the measurement of road traffic volumes in VKm
Luxembourg, 14 June 2007
Which method/s to use for collecting the data on VKm?
- Introduction
As presented in the UNECE Handbook on Statistics on Road Traffic, there are four main methods for the estimation of the volume of road traffic measured in VKm:
- Odometer readings at roadworthiness tests
- Household surveys
- Traffic counts
- Fuel consumption
This document describes briefly each method, presents their advantages and disadvantages and identifies possibilities for using them in a complementary or combined manner. Its main purpose is to stimulate discussion during the Task Force meeting on the choice of a suitable, pragmatic methodology for the collection of VKm data.
- Description of the methods – advantages and disadvantages of each one
2.1.Odometer readings from roadworthiness tests
This method relies on the odometer readings taken at roadworthiness tests which are mandatory in the EU Members States, according to Directive 96/96.
The calculations behind this method are the following. The odometer reading provides the kilometres driven by the tested vehicle in a specific period which are converted into kilometres driven per day. The period in question is either the interval between two successive tests or, lacking data from two tests, the interval between the current test and the first day of use of the vehicle.
The kilometres per day are multiplied by the number of registered vehicles to estimate the average daily volume of road traffic. Multiplying further by the number of days in a given year, gives the average yearly volume.
Errors may be introduced in the statistics due to a) errors in reading the odometer, b) the use of older odometers with 5-digits which reset when kilometres reach 99999, c) cheating with the odometer, d) changes in the use of vehicles between successive tests, e) unsuitability of the vehicle registers for statistical estimation.
Advantages of using odometer readings
Data can be broken down by vehicle characteristics to a very detailed level, since national vehicle registers contain detailed information about each vehicle’s characteristics.
The accuracy of statistics can be very high if data from all roadworthiness tests are recorded.
Disadvantages of using odometer readings
The traffic of foreign-owned vehicles within the national territory is not covered.
Vehicles that are not due for a roadworthiness test are not included.
As tests occur at irregular, different intervals per vehicle, the estimation of kilometres travelled in a specified reference period must rely on certain simplifying assumptions about yearly traffic or on data from third sources. Moreover, this irregularity means that accurate data cannot be produced too close to the end of the reference period.
Several countries do not record the measurements taken at odometer readings, as this is not mandated by the Directive.
2.2.Household surveys
This method relies on surveys of households, persons or vehicle owners, which measure, possibly together with other things, the use of vehicles and collect, inter alia, the number of kilometres travelled.
Such surveys are usually sample surveys. Since they are very often not specific vehicle use surveys, the target population, sampling frame, sample design, mode of administering the questionnaire may differ a lot between national surveys.
Advantages of household surveys
Additional information related to vehicle use, beyond kilometres travelled and vehicle characteristics can be collected. The nature of this information can be specified during the survey design phase.
Disadvantages of household surveys
The sampling frame might not be up to date or specifically geared to a vehicle use survey.
The costs of producing accurate statistics can be high.
The request for too detailed information may be burdensome for respondents, reducing its quality
In general, all issues affecting household surveys: sampling variation, non-response, coverage errors, recall errors, under-reporting by the respondents, etc.
2.3.Traffic counts
This method relies on regular censuses of road traffic intensity. ‘Census’ denotes that measurements are taken in all sections of motorways and other major roads and in a substantial sample of sections in regional roads.
The measurements in each section are traffic counts, i.e. counts of the number of vehicles passing a point established in the section, within a given period of time. The period of measurement consists of a small number of working days’ and holidays’ peak hours (usually approximately 10 days in total) spread among spring, summer and autumn. 'Four hours’ measurements are taken each day. Some countries use a smaller number of days in less ‘busy’ road sections. Others also measure a small sample of sections for a small number of night hours.
The counts in each section are converted to estimates of annual average daily traffic (number of vehicles passing each day on average from the counting point of the section). The latter is multiplied by the length of the section and the number of days in the year in order to estimate the total vehicle-kilometres. Where the counts come from a sample of section, it is assumed that not counted sections have the same traffic intensity as counted ones. Statistics can be broken down by vehicle characteristics.
The usual frequency of such censuses is five-yearly. Some countries complement them with measurements taken continuously in the most loaded sections by automatic counters.
Advantages of using traffic counts
All road traffic within the national territory is covered.
Disadvantages of using traffic counts
Due to resource constraints, traffic censuses are carried out at a multi-annual frequency.
2.4.Fuel consumption
This method aims to provide not only estimates of the volume of road traffic but also estimates of consumption per kilometre and total fuel consumption which are coherent between them and with official national data on fuel deliveries.
The method combines vehicle fleet size with estimates of average distance travelled by each vehicle (derived with a different method, e.g. a survey) and obtains an estimate of the volume of road traffic. The latter is combined with data on fuel consumption per kilometre (also derived from surveys or from vehicle reviews) and results in an estimate of total fuel consumption. This is then compared with total fuel deliveries, from which it should not differ excessively (more than an amount known as the Statistical Adjustment). If the difference is excessive the total fuel consumption figures are adjusted causing corresponding adjustments to the estimates of the volume of road traffic.
Statistics can be broken down by type of road or type of vehicle.
Advantages of fuel consumption
The method can be used as a control of road traffic estimates produced with a different method.
All road traffic within the national territory is covered.
Disadvantages of fuel consumption
The method needs statistics on the volume of road traffic obtained from an independent source (i.e. one of the three other methods).
- Complementarities of the methods – potential for combined use
The different methods of estimating the volume of road traffic have significant overlap in the types of traffic covered and the information they collect. Therefore, it is not practical to assign separate ‘targets’ to each one.
One exception, however, is that odometer readings could be combined with a sample survey of owners of new cars, provided that a reliable sampling frame is available for the latter. The separate survey’s costs would be justified since a large number of new cars are purchased each year and new cars are driven at longer distances than older ones contributing significantly to the total volume of road traffic.
On the other hand, the different methods can be combined in a national system for measuring road traffic. For instance, either odometer readings or a household survey can be used at regular intervals and censuses based on traffic counts can be used more infrequently as a benchmark. The latter can be planned to coincide with an implementation of the regular method so as to provide estimates to correct its possible bias.. The estimates can be used to adjust the regular method’s results in the years in between the next implementation of the traffic census counts.
The ‘fuel consumption’ method can also be used as a control of the regular estimates of volume of road traffic.
The choice between odometer readings and a household survey depends on a number of contradicting factors: more information can be collected with a survey but at greater cost and risk of errors, the survey can produce more timely results than odometer readings, etc. EU regulation on roadworthiness tests as well as established national statistical production practices will also play an important role in the decision.
1