1. 1 Previous Method – Use of a Survey

1. 
   General
   

1.1 Previous Method – Use of a Survey

Until 2011, the Kilometers Travelled Survey was based on reports received from the driver. The survey was conducted on sampling days at vehicle inspection stations, with the sampling days spread over the year of the survey. The annual estimate of kilometers travelled during the survey period was calculated as the average of the monthly reports of kilometers travelled reports, multiplied by 12. The number of vehicles was calculated as the average of the total number of vehicles at the ends of the last two years.

1.2 New Method – Use of Administrative Files

As of the 2011 data, which were included in the publication Kilometers Travelled 2012,¹ the data on kilometers travelled are based on administrative files obtained by the CBS from the Ministry of Transport.

Beginning in 2010, odometer readings were recorded during the annual inspection test of the vehicle. This information was recorded in the vehicle registry of the Ministry of Transport.

Calculation by the CBS of the annual kilometers travelled was based on a comparison of the odometer readings that were recorded in two different inspection tests of the same vehicle, at two different points of time, with a difference of approximately a year.

The number of vehicles was calculated according to the number of days the vehicle had a valid licence during the same period of time (See Chapter 4, Methodology, Paragraph 4.1, e.2).

The data obtained were thus actually for twelve moving months based on the date the vehicle passed the inspection test.

This new method enabled considerable expansion of the publication: adding distributions and detailed groupings that were not possible with the survey data, due to the size of the sample. For example, it was possible to differentiate among vehicles that were privately owned, leased vehicles, company vehicles, and rental vehicles, and to calculate the kilometers travelled according to the type of ownership. In addition, it was possible to present important characteristics in greater detail.

Tables 1–3 are summary tables that include data according to the old method, for 2011 and for the previous years. The data in Tables 4 and on are presented only according to the new method.

The findings refer to data from the years 2012–2013 obtained from processing of the administrative files.

In Israel in 2013, motor vehicles travelled a total of 51,207 million km. In comparison to 2012 (50,107 million km), this was an increase of 2.2%.

This increase resulted from a decrease of 0.5% in average annual kilometers travelled by vehicles, as well as from an increase of 2.7% in the number of active vehicles as compared to the previous year.

Most of the annual kilometers travelled were by private cars – 74%. The proportion of trucks was 17.8% and that of taxis was 3.2%.

Examination of the changes that occurred since 2000 showed that the annual kilometers travelled increased by 40% and the total number of active vehicles increased by 56% compared to the year 2000. However, the area of the roads increased by 35%, and the length of the roads increased by 14%.

In 2013 the average annual kilometers travelled by private cars was 16.4 thousand km (as in 2012). The total kilometers travelled by all private cars in this year was estimated at 37,848 million km, constituting 74% of the total annual kilometers travelled by all vehicles in this year. In comparison to 2012 (36,545 million km), this was an increase of approximately 3.6%.

This increase resulted entirely from an increase of 3.6% in the number of active private cars as compared to the previous year, whereas the average annual kilometers travelled by private cars remained almost unchanged.

Private cars by engine capacity: In 2013, the average annual kilometers travelled by private cars with an engine capacity of up to 1,000 cc was 14.8 thousand km. The average annual kilometers travelled continued to increase, to 16.9 thousand km for vehicles with an engine capacity of 2,001 to 4,000 cc. (Vehicles with an engine capacity of more than 4,001 cc were not included due to the high average age of these vehicles. See Table 16 in the publication Motor Vehicles 2013²).

Private cars owned by leasing companies: In 2013, the average annual kilometers travelled by private cars owned by leasing companies was 31.2 thousand km, as compared to 29.8 thousand km travelled by private rental or touring cars, 27.7 thousand km by private company cars, and only 14.4 thousand km by privately owned private cars.

Private cars fuelled by diesel oil or gas: In 2013, private cars fuelled by diesel oil travelled an average of 25.3 thousand km and private cars fuelled by gas travelled 24.1 thousand km, compared to 16.1 thousand km travelled on average by private cars fuelled by gasoline.

2.3 Trucks

In 2013, the average annual kilometers travelled by trucks (gasoline and diesel) was 26.8 thousand km (as in 2012). The total kilometers travelled by trucks was 9,099 million km, 17.8% of the total annual kilometers travelled by all vehicles in this year. In comparison to 2012 (9,347 million km), this was a decrease of 2.7% that was caused by a decrease of 2.4% in the number of active trucks, while the average annual kilometers travelled by trucks remained the same.

The average annual kilometers travelled by diesel trucks in 2013 was 28.8 thousand km, twice the amount travelled by gasoline trucks (14.4 thousand km).

Trucks by gross weight: As the gross weight of the trucks increased, their average annual kilometers travelled increased. In 2013, the average annual kilometers travelled by diesel trucks with a gross weight of 34 tons and up was 71.2 thousand km, which was more than twice the average kilometers travelled by diesel trucks with a gross weight of up to 3.6–12.0 tons (only 33.8 thousand km).

2.4 Buses

In 2013, the average annual kilometers travelled by buses was 58.8 thousand km (59.3 thousand km in 2012). The total kilometers travelled by buses was 950 million km (918 million km in 2012). The number of buses increased in 2013 by approximately 4.4% as compared to the previous year.

The total annual kilometers travelled by buses constituted 1.9% of the total annual kilometers travelled by all vehicles in 2013 (51,207 million km).

2.5 Minibuses

In 2013, the average annual kilometers travelled by minibuses was 50.5 thousand km (49.8 thousand km in 2012). The total kilometers travelled by minibuses was 716 million km (719 million km in 2012). The number of minibuses decreased in 2013 by approximately 1.8% as compared to 2012.

The average annual kilometers travelled by public minibuses in 2013 was 63.4 thousand km, compared to the average annual kilometers travelled by private minibuses – only 39.3 thousand km.

The total annual kilometers travelled by minibuses in 2013 constituted 1.4% of the total annual kilometers travelled by all vehicles in this year.

In 2013, the average annual kilometers travelled by motorcycles was 7.2 thousand km (7.1 thousand km in 2012). The total number of kilometers travelled by motorcycles was 872 million km (834 million km in 2012), which constituted 1.7% of the total annual kilometers travelled by all vehicles in this year.

In 2013, the average annual kilometers travelled by taxis was 78.2 thousand km (79.5 thousand km in 2012). The total kilometers travelled by taxis was 1,640 million km (1,661 million km in 2012), constituting 3.2% of the total annual kilometers travelled by all vehicles in this year.

According to the provisions of that licence, a bus can carry seated as well as standing passengers. Excludes a minibus, which is a unique type of bus existing in Israel, licenced to carry up to 17 persons (including the driver).

Up to 1992, a minibus was included under "Truck up to 4 tons - van" (i.e., without separation between the passengers and the back load).

Two main types of trucks are distinguished by gross weight:

The population investigated included all motor vehicles registered at the Ministry of Transport with a valid licence or a licence that expired during the year of the investigation (the active vehicles).

It did not include army and police vehicles, hauled vehicles, tractors, vehicles of foreign citizens and of holders of provisional vehicle licences, vehicles registered in the Palestinian Authority, vehicles of tourists staying in Israel less than three months, diplomatic and U.N. vehicles, etc.

b. The Investigation Period

The period investigated was the year of the publication.

Administrative files regarding motor vehicles received from the Ministry of Transport.

The data that appear:

• 
  Information on characteristics of the vehicles such as: year of production or registration, engine capacity, gross weight of trucks.
  
• 
  Information on vehicle ownership: type of ownership, owner’s locality of residence.
  
• 
  Information from the inspection stations: date of inspection test, cumulative kilometers travelled at time of inspection test.
  

At the CBS, each vehicle file underwent a process of intake and determination that the vehicle was active (the file received was a historical file that began at the establishment of the State of Israel, meaning that no vehicle records were deleted).

To calculate annual estimates, the CBS used cumulative data of kilometers travelled from two different dates, with a difference of about a year between the two. In this manner, the kilometers travelled data that was received was for twelve moving months, according to the timing of the vehicles' inspection test. Part of the kilometers travelled data that were reported included a period of time not included in the year investigated, and another period where part of the year was not covered.

The calculation of the annual kilometers travelled by vehicles was based on the difference between the cumulative kilometers travelled reports of two consecutive inspection tests.

In order to enable optimum calculation for a specific calendar year, the largest number of relevant months was taken into account for each vehicle, according to the date it underwent its inspection test. Therefore, for a specific calendar year under investigation, the data that were used were those from June of the previous year to June of the year following the investigation year. In practice, for vehicles undergoing an inspection test during the first half of the year, the data taken into account were for the survey year and the following year. For vehicles undergoing an inspection test during the second half of the year, the data taken into account were for the survey year and the previous year. Thus, symmetry was achieved in solving the problem of coverage of the calendar year.

For example, the 2012 data were calculated as follows:

For a vehicle that underwent an inspection test during January, the January 2012 and January 2013 data were used: 11 months from the survey year and an additional month from the following year.

For a vehicle that underwent an inspection test during June, the June 2012 and June 2013 data were used: 6 months from the survey year and an additional 6 months from the following year.

For a vehicle that underwent an inspection test during September, the September 2012 and September 2011 data were used: 9 months from the survey year and 3 months from the previous year.

Since the two dates of the inspection tests were not necessarily exactly one year apart, the difference between the two reports was standardized according to the number of days that passed between the two reports so that they fit one year. Furthermore, some of the vehicles were not active the entire year. The most notable example was new vehicles that were registered at some point during the year, and thus were not active the entire year. Another example was vehicles that were active at the beginning of the year but during the year became inactive (for example, vehicles that were scrapped or stolen). For this purpose, a calculation was performed to determine the number of days each vehicle was active during the investigation period (number of active days during the year), and the kilometers travelled was calculated according to this number. Thus, the annual kilometers travelled by vehicles was calculated according to the difference between the cumulative kilometers travelled divided by the difference in days that passed between the two reports, multiplied by the number of days during the year that the vehicle was active (with a valid licence). This calculation method included the greatest number of relevant months in the year investigated.

Because the calculation of kilometers travelled by vehicles was based on the odometer readings from two separate occasions, the quality of both of these readings was of great importance. Therefore, the readings underwent a process of logic checking, whose purpose was to detect vehicles with missing or atypical odometer recordings that most likely did not reflect the actual kilometers travelled and therefore might have affected the estimates. The logic checks were adapted to each type of vehicle separately according to its different characteristics. The checks were conducted on each individual report of kilometers travelled that was included in the calculation for annual kilometers travelled, as well as for the difference between the two readings relative to the number of days between the two dates. In the case of a vehicle for which the kilometers travelled report(s), or the difference between the two reports, was atypical (high or low in an atypical way), the annual kilometers travelled was imputed by the CBS.

Imputation, which is the estimation of a missing or disqualified datum, was used in order to minimize estimation biases resulting from non-receipt or disqualification of the required reports.

For some vehicles, reports (readings) were not received for the two relevant occasions. Some of these vehicles were supposed to have no reports, and some were only supposed to have one report. For example, new private cars are required to undergo an inspection test only two years after being registered for the first time. For some of the vehicles, the number of kilometers travelled was not reported, even though this was required. For a small portion of the vehicles having two reports, the kilometers travelled datum was disqualified due to failing the logic checking criteria, including cases in which the report for the current year showed cumulative kilometers travelled lower than the cumulative kilometers travelled for the previous year. The number of annual kilometers travelled was imputed for all vehicles that did not have two consecutive, valid reports for the investigation year.

Two methods of imputation were used for annual kilometers travelled:

1. 
   Imputation of kilometers travelled for a vehicle, based on the report of kilometers travelled for a long period of time.
   
2. 
   Imputation of homogeneous groups.
   

In the second imputation method, the vehicles were divided into groups that were homogeneous with respect to the investigated variable, using a series of auxiliary variables. The major variable according to which the imputation groups were formed was the main type of vehicle, within which sub-groups were formed according to different characteristics. Thus, for example, within the group of private cars, the sub-groups were formed according to a combination of type of vehicle ownership (private/company), age group, and engine capacity group. For trucks, the sub-groups were formed according to type of fuel (gasoline/diesel oil), vehicle ownership, age group, and gross weight. In addition, in some of the groups, there was a division into two types of trucks.

Using this method, each vehicle that was missing a value for annual kilometers travelled was imputed the average annual kilometers travelled in a homogeneous group that was calculated based on all vehicles that had reports on two dates and that passed the logic checking.

f. Reliability of the Data

The data for the publication in the new format were not based on a sample and were therefore not subject to sampling errors. However, as in all statistical work, the data were subject to non-sampling errors.

The population investigated was identical to the population in the new method. See paragraph 4.1.a.

The investigation period was identical to the period of the new method. See paragraph 4.1.b.

An annual, dedicated survey conducted by the CBS at the inspection stations, as well as information obtained from bus companies owning licenced garages.

The sampling included all the inspection stations that licenced civilian vehicles of the general population, not including small inspection stations.

The sample included 4 quarter-year samples. Each quarter-year sample was composed of the following samples:

1. 
   Sample of small stations – each small station was enumerated on one day during the quarter-year.
   
2. 
   Sample of days at other stations (the vast majority of the sample) – each station was enumerated on 1–4 days of the quarter-year.
   

The sample did not represent:

1. 
   Buses owned by large transportation companies (approximately half of the buses), which were licensed in a licensed garage under their ownership. The representation of these buses was based on complete information received directly from the companies.
   
2. 
   New private and commercial vehicles, which were exempt from the annual registration test at the end of the vehicle's first year of activity. The estimates for the first year of these vehicles were imputed according to the data received regarding new vehicles that underwent an inspection test at the end of their second year (a sample test done in the past showed that this imputation is reliable).
   

Every vehicle owner was asked how many kilometers were travelled in the past month and in the past 12 months. The owner was also asked to state the type of fuel used in the vehicle, the vehicle’s fuel consumption (km per liter), and the expenditure on fuel in the past week.

Because the data were based on a survey, they were subject to sampling errors and non-sampling errors.