Fuel Consumption

Detailed information for December 1988

The Fuel Consumption Survey was designed to provide a data base that would allow for the monitoring of average fuel consumption in personal-use passenger cars, light trucks and vans.

Data release - -

• Questionnaire(s) and reporting guide(s)
• Description
• Data sources and methodology
• Data accuracy

Description

The Fuel Consumption Survey came into being as a result of concerns about energy use in the late 1970's. The federal government of the time was concerned about reducing the use of petroleum fuels and, in particular, motor vehicle fuel consumption. In 1977, Transport Canada made a proposal to Statistics Canada to conduct a fuel consumption survey. The purpose of the survey was to establish a data base concerning personal-use of passenger cars, light trucks and vans in Canada. The data base would contain information on:
- number of personal-use passenger cars;
- total distance travelled;
- average fuel consumption per unit of distance travelled; and
- seasonal fluctuations in fuel consumption and in distance travelled.

Data on fuel consumption patterns were collected from households by asking respondents to keep a diary of their mileage and fuel purchases for one month. Data were collected for all twelve months of the year. Data collection for the Fuel Consumption Survey commenced in October 1979 and ended in December 1988.

Subjects

• Energy
• Energy consumption and disposition

Data sources and methodology

Target population

When the Fuel Consumption Survey commenced in 1979, the target population of the survey consisted of personal-use passenger cars operated in Canada. Between October 1981 and December 1987, the survey was expanded to include personal-use light trucks and vans operated in Canada. In 1988, the target population was reduced to include only personal-use passenger cars. The target population does not include rental vehicles, taxis, fleet operated vehicles, driver-training vehicles and antique vehicles.

Sampling

This is a sample survey with a cross-sectional design.

Samples for the survey were selected from sample frames created using the provincial motor vehicle registration files, which were provided to Statistics Canada every three months. The registration files contained a variety of auxiliary information including vehicle descriptions (make, model, model year), owner's name and address, vehicle weight, body style, vehicle use, vehicle identification number and license plate number. This information was used to exclude vehicles that were identified as not being part of the target population, for example, buses, trailers, motorcycles, rental vehicles, government-owned vehicles and vehicles not licensed for the current year.

Given that motor vehicle registration procedures differ from province to province, and in order to produce national estimates, there is a requirement that sampling populations should be consistent across provinces. Therefore, screening was also incorporated in the data collection procedures to ensure that data were collected for the same classes of vehicles in all provinces.

For purposes of sample selection, the vehicles on the registration files served as sampling units. The principal driver of the vehicle is the responding unit, while the unit of reference as well as the unit of analysis is the personal use vehicle.

The basic design of the Fuel Consumption Survey constitutes a stratified systematic sample design. In order to improve the efficiency of the sample design the frames were stratified by model year and vehicle weight (wheelbase or number of cylinders were used if weight were not available) for passenger cars, and by model year for light trucks and vans. It should be noted that the number of strata varied over time.

A stratified systematic sample was selected quarterly for each province and was screened for ineligible vehicles. Systematic samples were then selected monthly from these larger samples. For each strata efforts were made to sample a minimum number of cars. The strata "vehicle weight" had new boundaries set each year to reflect the new population.

Data sources

Responding to this survey is voluntary.

Data are collected directly from survey respondents.

The method of data collection employed a combination of a telephone interview and a mail-out/mail-back Fuel Purchase Diary. Respondents were initially contacted by telephone and asked some screening questions concerning vehicle type and use. Ineligible vehicles (e.g., vehicles not operated during the survey month, vehicles not used for personal use, scrapped vehicles), encountered at this stage were dropped from the sample. Fuel purchase diaries were mailed for each car included in the screened sample. The principal driver of each car was asked to ensure all fuel purchases were recorded for the survey month. The date of purchase, odometer reading at the time of each purchase, quantity of fuel purchased and the total amount paid, price per unit of fuel, type of fuel were all to be recorded.

Until December 1985, the number of diary mail-outs per month was approximately 1,000 diaries for passenger cars and 800 diaries for vans throughout Canada. These numbers were cut between January 1986, through December 1987, to approximately 333 passenger cars and 260 light trucks and vans; in 1988, only 333 passenger cars were sampled monthly.

These samples were allocated proportionally to the population size of each province, subject to minimum sample requirements for each strata. Since the number of telephone interviews conducted each month varied depending on the number of telephone non-interviews and on the number of out-of-scope vehicles encountered, sufficient numbers of vehicles had to be selected each month to ensure the diary mail-out requirements were met.

View the Questionnaire(s) and reporting guide(s).

Error detection

To maximize the accuracy of the data, quality control procedures were applied throughout the collection and processing of the diaries. At the beginning of the survey month, phone calls were made to each principal driver to ensure that they received the diary and to answer any questions they may have had. At the end of the month, diaries were returned to Statistics Canada. To improve the response rate, follow-up phone calls were made for diaries not returned. All the returned diaries were screened manually before computer processing. Diaries with less than two purchases or diaries on substituted vehicles were rejected at this phase of the survey. The remaining diaries were captured and edited. Editing was carried out in stages with diaries passing through a total of five edit stages.

Imputation

Incomplete diaries were imputed. For example, to obtain estimates on fuel consumed and distance driven, the diaries that were incomplete for part of the survey month were adjusted upwards to reflect the complete month, unless it was known that the vehicle was not operated for part of the month.

Estimation

The estimation of population characteristics from a survey sample is based on the premise that each sampled unit represents a certain number of unsampled units in the population. A weight is attached to each unit to reflect this representation. For the Fuel Consumption Survey a unit is a personal-use automobile. Weights for the sample were obtained by using provincial Motor Vehicle Registration Files to determine control totals. Control totals were adjusted to reflect only the eligible population in the survey.

The weighting procedure commenced once a clean survey data file had been created. The records for valid respondents were then weighted to produce estimates of the target population. The weighting procedure accounts for non-response to the screening questionnaires and diaries. It also adjusted for out-of-scope vehicles in the original sample but not belonging to the target population.

The estimates were derived on the assumption that the characteristics of the non-respondents do not differ from those of the respondents. However, if non-respondents differ from respondents on automobile usage characteristics, then biased estimates will have resulted. In order to decrease the effect of possible non-response bias, the sampling weight adjustments were conducted at the stratum level.

Non-response could occur at two phases in this survey: at the telephone screening phase and at the diary phase. Non-response at the telephone phase was primarily due to an untraceable telephone number for the registered owner. Other reasons were no answer after several calls, refusal, temporarily absent and moved out of province. Non-response at the diary phase could occur because the driver forgot to keep the diary, because the diary was not mailed back, or because the information was not useable. In order to compensate for non-response, vehicles for which responses were obtained must represent more vehicles than intended in the survey design; thus their sampling weights were increased.

Disclosure control

Statistics Canada is prohibited by law from releasing any information it collects that could identify any person, business, or organization, unless consent has been given by the respondent or as permitted by the Statistics Act. Various confidentiality rules are applied to all data that are released or published to prevent the publication or disclosure of any information deemed confidential. If necessary, data are suppressed to prevent direct or residual disclosure of identifiable data.

In order to prevent any data disclosure, confidentiality analysis is done using the Statistics Canada Generalized Disclosure Control System (G-Confid). G-Confid is used for primary suppression (direct disclosure) as well as for secondary suppression (residual disclosure). Direct disclosure occurs when the value in a tabulation cell is composed of or dominated by few enterprises while residual disclosure occurs when confidential information can be derived indirectly by piecing together information from different sources or data series.

Revisions and seasonal adjustment

This methodology does not apply to this survey.

Data accuracy

October to December 1979:
For this quarter, October to December 1979, telephone interviews were conducted with over 3,465 vehicle owners and 2,825 diaries were mailed. The diary return rate for this quarter was 64.4%. The useable diary return rate was 59.5%.

October to December 1988:
During the telephone screening phase, interviews were attempted for 1,751 vehicle owners. Completed interviews were obtained with 1,170 owners for a response rate of 66.8%. Of those respondents to the telephone screening, 1,007 were eligible to take part in the survey, a rate of 86.1%. Of those eligible, useable diaries were returned by 562, a response rate of 55.8% to this phase of the survey.