EXPLANATORY NOTES


MOTOR VEHICLE CENSUS

1 This publication presents statistics relating to vehicles which were registered with a motor vehicle registration authority at 31 March 2008.

2 Motor Vehicle Censuses (MVC) have been conducted periodically from 1971 on each of the following dates: 31 March 2008, 31 March 2007, 31 March 2006, 31 March 2005, 31 March 2004, 31 March 2003, 31 March 2002, 31 March 2001, 31 October 1999, 31 October 1998, 31 October 1997, 31 October 1996, 31 May 1995, 30 June 1993, 30 September 1991, 30 September 1988, 30 September 1985, 30 September 1982, 30 September 1979, 30 September 1976, 30 September 1971.


SCOPE

3 The Motor Vehicle Census includes all vehicles registered with a state, territory or other government motor vehicle registry for unrestricted use on public roads with the following exceptions:

• recreational vehicles such as trail bikes and sand dune buggies intended for off-road use in most states and territories (in Victoria and Queensland these vehicles must be registered and are thus included in the statistics);
• consular vehicles; and
• vehicles registered by the defence forces.

4 Vehicles on register are defined as those vehicles registered at the date of the census, or had registration expire less than one month before that date.

5 Motor vehicle registration statistics are derived from data made available by various state and territory motor vehicle registration authorities and reflect the information as recorded in registration documents.

6 Vintage and veteran vehicles are excluded from vehicle counts in the published data, but are available for consultancy data requests. These vehicles are identified using a unique insurance code, with the following exceptions:

• in New South Wales this insurance code is not provided and therefore it is not possible to separately identify these vehicles. Therefore veteran and vintage vehicles are included in the number of registered vehicles for New South Wales.
• while Victoria does provide an insurance code for these vehicles, this insurance code also applies to tractors, motorcycles and self propelled farm machinery owned by primary producers and used solely in connection with the business of a primary producer. These records are deemed out of scope. (Note: tractors and motor cycles not used solely in connection with the business of a primary producer have a different insurance code which deems them in scope.) Therefore, the number of veteran and vintage vehicles cannot be obtained for Victoria.

7 The estimated proportion of motor vehicles, excluding motor cycles, taken off the register since the previous MVC is referred to as the attrition rate. The attrition rates in the following table have been calculated using new motor vehicle sales data.

ESTIMATED ATTRITION(a), State/territory - Census years
	Between 2003 and 2004		Between 2004 and 2005		Between 2005 and 2006		Between 2006 and 2007		Between 2007 and 2008
	no.	%	no.	%	no.	%	no.	%	no.	%
NSW	196 609	4.7	207 506	4.9	214 431	4.9	217 518	4.9	178 856	3.9
Vic.	179 738	4.9	173 120	4.7	171 489	4.5	179 505	4.6	178 750	4.5
Qld	88 012	3.3	102 952	3.7	98 452	3.4	97 337	3.2	107 629	3.4
SA	44 813	4.0	48 779	4.3	40 073	3.5	45 253	3.9	43 554	3.7
WA	46 330	3.1	46 716	3.1	38 310	2.4	46 866	2.8	61 171	3.5
Tas.	6 768	1.9	8 961	2.5	8 272	2.2	13 258	3.5	10 679	2.7
NT	6 320	5.8	5 390	4.8	5 925	5.1	5 649	4.7	5 983	4.8
ACT	r8 346	r3.8	r11 207	r4.9	r10 799	r4.7	r10 227	r4.3	10 554	4.4
Aust.	r576 936	4.2	r604 139	4.3	r587 751	4.1	r615 613	4.1	597 176	3.9
r revised
(a) These estimates exclude motorcycles. For more details on the calculation of attrition rate refer to glossary.

METHODOLOGY

8 The 2008 Motor Vehicle Census is a count of all vehicles registered with a state or territory motor vehicle registration authority (MVR) at 31 March 2008. The snapshot date has consistently been 31 March since 2001. The Australian Bureau of Statistics (ABS) receives the data from the MVRs. At this stage all records are compared to the previous MVC data for vehicles that match (on-going registrations) and relevant details are updated. The matching is based primarily on a unique identifier given by the registries to each record. After the primary match is done, a secondary match is run based on the Vehicle Identification Number (VIN), engine number, chassis number, year of manufacture and body to improve the number of matched records.

9 The un-matched records (new vehicle registrations) undergo comprehensive editing. This editing checks such items as body, make, model, year of manufacture, weight, heavy vehicle code, postcode and fuel.

10 The editing of on-going registrations is limited as these records have been edited in past years. Finally, aggregate data is checked across time, states and vehicle types to ensure coherency.


DATA QUALITY

11 The size of the registered motor vehicle fleet (over 15 million vehicles) dictates that quality assurance of each record is not possible. Considerable effort is made, both by motor vehicle registration authorities and the ABS, to improve data quality. This has been done through the introduction of improved processing and classification systems and the incorporation of additional internal checking to ensure the vehicle counts track consistently in a time series when compared to previous MVCs. While these have improved data accuracy, there remains some variation in the reporting from different state and territory registries and hence care should still be taken when comparing data across jurisdictions. The ABS recognises the importance of motor vehicle statistics to a large variety of statistical users. While the data are made available at a detailed level, the quality of this detailed data is less certain, therefore the ABS is unable to guarantee that it is necessarily sufficient for all purposes for which it is used.

12 For vehicles manufactured during or after 1990, the statistics in this publication are based on procedures using the VIN, adopted by state and territory motor vehicle registration authorities. The VIN system allows accurate identification of the make and model of a vehicle, which can then be used to better identify the vehicle type. Vehicles manufactured earlier than 1990 generally do not have a VIN. For these earlier vehicles, standard ABS makes and models were produced from registry make and model data using information such as engine number and chassis number.

13 Each year, for new vehicle registrations, the Registry's vehicle model description is converted to a standard ABS model. This is necessary as model descriptions can vary between Registries. Resource constraints affect the amount of model conversion that can be carried out, resulting in some vehicles being left with a blank model which is then coded to 'model unknown'.


Data confrontation

14 While there are no data sources available in which to directly assess the quality of MVC data, new motor vehicle sales data and population data provide an indication of data quality as they give some context to the size of the motor vehicle fleet. Assessing the quality of data by comparison against other datasets is known as data confrontation. The ABS has been using sales of new motor vehicles data and estimated resident population (ERP) data to 'confront' MVC data for several years.

New Motor Vehicle Sales (NMVS)

15 Sales of new motor vehicles can be compared with data from the MVC based upon the year of manufacture. The following assumptions have been made:

• vehicles manufactured in one year are not available for sale until after 31 March of that year;
• vehicles manufactured in one year are sold as new vehicles until at least 31 March of the following year;
• vehicles sold over recent years will remain on the MVC year after year.

16 Therefore, it can be assumed that vehicles manufactured in 2007 (as reflected in 2008 MVC data) would most likely have been sold in the period between 1 April 2007 and 31 March 2008 (as captured by sales of new motor vehicles data). Similarly, vehicles manufactured in 2006 would have been sold between 1 April 2006 and 31 March 2007, and so on.

17 The graph below shows the relationship between sales of new motor vehicles and MVC registrations of vehicles manufactured in the year preceding the snapshot date.

Estimated Resident Population (ERP)

18 The table below shows a comparison of growth on a state/territory basis for both the MVC and ERP. Over the past few years, strong motor vehicle sales have meant that the growth in motor vehicle registrations has been higher than the growth in ERP. However, growth in MVC and ERP has tended to be in the same direction.


STATISTICAL GEOGRAPHY

19 The only geographic data available from the MVC are postcode of owner and state or territory of registration.

20 If analysing regional data based on postcodes, the following considerations need to be taken into account:

• there are some instances where the postcode of the vehicle's owner is in a different state or territory from that in which the vehicle was registered. For example, the postcode of the owner of the vehicle could be in Sydney, yet the vehicle could be registered in Qld.
• some official postcodes (P.O. boxes, competition postcodes) do not correspond to residential areas
• some postcodes could be invalid.

RELATED PUBLICATIONS AND PRODUCTS

21 Users may also wish to refer to the following publications and products which contain information relating to motor vehicles in Australia:

• Sales of New Motor Vehicles, Australia, (Electronic Publication) cat. no. 9314.0 (issued monthly)
• Survey of Motor Vehicle Use, Australia cat. no. 9208.0 (issued annually since 1998) or 9210.0.55.001 (issued annually)

ABS DATA AVAILABLE ON REQUEST

22 As well as the statistics included in this publication, the ABS has other relevant data available on request. Inquiries should be made to the National Information and Referral Service on 1300 135 070.