The Survey of Motor Vehicle Use (SMVU) is conducted by the Australian Bureau of Statistics (ABS) to provide specific information on the use of motor vehicles in Australia, suitable to the needs of the transport industry, private industry generally and government.
In response to some ongoing user concerns with results from the SMVU, the ABS is conducting a 'Fitness for Purpose' review of the SMVU. The objective of the review is to assess the extent to which the data currently collected by the SMVU meets users' needs, i.e. the right data items with the right statistical attributes (for example: reliability, frequency of data collection, level of geographic dissaggregation, level of cross-classification possible), and investigate alternative options for better meeting users' needs. The review Terms of Reference (cat. no. 9208.0.55.001) provide more details on this review.
The review is being conducted in three stages. Phase 1 of the review involved extensive consultation with users regarding their needs, priorities and concerns with the SMVU. This report is the final output from phase 1. Phase 2 of this review will involve identifying possible options to address the issues outlined in this document, then prioritising these options and producing the final review recommendations. A summary of the recommended phase 2 investigations appears in section 1.6. The final review recommendations from phase 2 will be implemented in phase 3.
1.1 Purpose of this paper
1.3 Objectives of the review
1.4 SMVU resourcing
1.5 Summary of the consultation process
1.6 Summary of recommended investigations for phase 2
2.0 Data issues this review will address
2.1 Authoritative data set
2.2 Data volatility
2.3 Measuring level movements in key variables
2.4 Geographic patterns of activity
2.5 Data frequency
2.6 Quality of fuel use estimates
2.7 Impact of unregistered vehicles and out of scope vehicles
2.8 Vehicle classification used by the ABS
2.9 Question changes on SMVU survey forms
3.0 Data issues this review will not address
3.1 Capturing transport activity by industry
3.2 Cross sector use of transport data: environmental impacts
3.3 Cross sector use of transport data: financial performance of the transport industry
Appendix A. Data quality framework
Appendix B. Cross reference with data issues from the Transport Information Development Plan
Appendix C. Wider transport issues out of scope for this SMVU review
Appendix D. SMVU 2003 achieved relative standard errors
Appendix E. AustRoads and ABS vehicle classifications
Review Terms of Reference
1.1 Purpose of this paper
This document is the final output from phase 1 of the SMVU Fitness for Purpose review. It reports the views of users, expressed during user consultation, regarding users' needs, priorities and concerns with the SMVU. It also identifies how SMVU data is currently used. This allows the SMVU review team to start phase 2 with a clear understanding of the issues to be addressed.
The paper outlines a range of issues. The issues are separated according to those deemed within scope of the review, to be investigated during phase 2 of the review, and those deemed beyond the scope of the current review. A summary of these issues is provided in section 1.6. This paper also recommends investigations to start determining options to address the issues. Please note that these investigations may be modified when the review team begins preparing for phase 2.
In undertaking phase 1 of the review, the review team used the ABS data quality framework to help identify the 'fit for purpose' nature of the SMVU data. A summary of this framework is provided in Appendix A for reference.
This review is an outcome of the Transport Information Development Plan (IDP) process conducted jointly by the ABS and the Transport Statistics Users Group (TSUG) in July 2003. The IDP process involved the ABS and TSUG working together to define and document key transport issues and the associated data needs of transport data users. In addition to defining key data needs, the Transport IDP identifies the role of various agencies as data suppliers and outlines strategies for cost effectively addressing the various data needs.
The Transport IDP process discussed a range of issues related to the SMVU. These form the basis of this review.
Appendix B provides a summary of the links between the data issues discussed in this paper and those issues outlined in the Transport IDP process in July 2003 and documented in the Transport IDP. The Transport IDP also covered a range of issues outside the scope of the SMVU and this review. Those issues are listed in Appendix C.
This review will address only issues relevant to the SMVU. All wider transport issues that are raised or discussed with users during this review will be progressed separately by ABS staff and external organisations.
1.3 Objectives of the review
The objective of this review is to critically assess the SMVU to ensure survey data aligns with users' needs, ie. the right data items with the right statistical attributes (for example: reliability, frequency of data collection, level of geographic dissaggregation possible, level of cross classification possible).
This review will achieve the following outcomes:
A. A methodology for SMVU which is the most appropriate to meet users' needs given the constraints under which the survey is conducted. The review will either recommend enhancements or conclude that the existing methodology is appropriate.
B. All parties will have a better understanding of how various stakeholders use the SMVU data to inform policy decisions.
C. All parties will have a better understanding of exactly what the SMVU is designed to measure and how it fits in with other data sources.
With regard to outcome A, section 1.6 summarises the investigations that will be conducted to seek to address the range of issues identified through phase 1 of this review.
The output of phase 2 may include recommendations that cover changes to the survey methodology. The impact of all recommended changes will be investigated during phase 2 of the review and documented in the phase 2 report. ABS responsibilities and policies relating to provider load and form design may impact on the feasibility of some options.
All solutions implemented in phase 3 must be covered by the existing survey budget and this may constrain which solutions can be implemented. Any trade-offs with regard to accuracy, data coverage, frequency, etc., will be documented in the phase 2 report. The ABS will consult with stakeholders, through TSUG, prior to the implementation of any changes to the SMVU. The ABS will attempt to maintain time series continuity where possible. Any trade-off between improvements in the survey accuracy and reliability and time series continuity will be considered very carefully.
The ABS will work with stakeholders to prioritise solutions within the current budget constraints. The range of feasible solutions to enhance the data quality may be expanded if a degree of user funding becomes available.
1.4 SMVU resourcing
The ABS undertakes the SMVU to provide data for the transport industry. The SMVU does not provide any data used by the ABS for its other published statistics such as the National Accounts. Within the current resources of the ABS, no additional funding is available for the SMVU. Consequently, any enhancements to the SMVU must be within the current SMVU budget, unless some additional funding is provided by transport data users.
The ABS would welcome discussions with stakeholders regarding any possibilities for user funding. Some other ABS surveys, such as the Survey of Tourist Accommodation, are enhanced by additional funding provided by the data users.
1.5 Summary of the consultation process
The SMVU review team began phase 1 of the review by releasing a discussion paper in October 2004. This paper documented known issues relating to the SMVU. Initial information on these data issues was obtained from the revised Transport IDP, discussed at the TSUG meeting in July 2003. Prior to releasing the discussion paper, the SMVU review team consulted with a small number of significant users to ensure that all the significant issues had been identified in the paper.
The discussion paper was sent to a range of major SMVU data users to obtain their feedback. Meetings were held with some stakeholders to discuss their issues with the SMVU. Other stakeholders provided written feedback and were contacted for further clarification if required.
Due to time constraints it was not possible to consult with all users of the SMVU. However, based on the feedback from those organisations that were contacted, the review team is confident that the range of issues covered in this document are an accurate and comprehensive statement of the range of user issues with the SMVU.
The review team thanks all stakeholders who provided feedback. That feedback obtained forms the basis of this report.
The stakeholders invited to participate in the consultation process are listed below. These organisations were either directly contacted by the ABS, or through AustRoads and/or the National Transport Data Framework Implementation Group. Approximately three quarters of these organisations provided feedback, either written, oral, or both.
- Bureau of Transport and Regional Economics
- NSW Department of Infrastructure, Planning and Natural Resources
- SA Department of Transport and Urban Planning
- Victorian Department of Infrastructure
- Queensland Department of Transport
- Queensland Office of Economic and Statistical Research
- Queensland Department of Main Roads
- Apelbaum Consulting Group Pty Ltd
- ACT Department of Urban Services
- NT Department of Infrastructure, Planning and the Environment
- ARRB Transport Research Pty Ltd
- NSW Roads and Traffic Authority
- Australian Local Government Association
- Main Roads WA
- WA Department for Planning and Infrastructure
- Tasmanian Department of Infrastructure, Energy and Resources
- National Transport Commission
1.6 Summary of recommended investigations for phase 2
Arising from the feedback obtained through Phase 1 of the review, the review team recommends nine investigations are conducted to seek to address the issues outlined in section 2 of this document. This section contains a summary of these investigations recommended for phase 2. These investigations are discussed in more detail in their respective issues sections of this document. The investigations may be modified as Phase 2 commences.
Please note that these are not the final recommendations of the SMVU review. Final recommendations will be developed from the results of the investigations and presented to the review project board for consideration.
As already mentioned, the results of the investigations conducted in Phase 2 of the review may lead to recommendations for changes to the SMVU. The review team is aiming to complete the high priority investigations during 2005. The sample for the 2005-06 SMVU is selected in August 2005, so it will only be possible to implement minor changes in time for the 2005-06 SMVU. More significant changes arising from the review will be implemented in the 2006-07 SMVU. The ABS will consult with the TSUG before making any changes to the survey.
Resource constraints may impact on the completion of all the recommended Phase 2 investigations in 2005. The review team has therefore ranked the issues in order of priority. It is intended that the issues for Phase 2 of the study will be undertaken in priority order.
A summary of the recommended investigations is listed below in (descending) order of priority. As some issues are closely related some investigations will address more than one issue. The recommendations, which are discussed in more detail in section 2 of this report, are grouped below where appropriate. Table 1.6.1 provides a summary of the investigations and the associated issues that are discussed in section 2.
1. Investigate disparities between SMVU data and other industry measures
- Joint ABS/Bureau of Transport and Regional Economics (BTRE) analytical project investigating disparities between SMVU data and industry measures, including fuel use data. This investigation will use SMVU unit record data. (issue 2.1)
- Consider options for publishing some data confrontation analyses where the SMVU data is compared over time and with other sources. (issue 2.1)
2. Investigate options for reducing the sampling error of the survey estimates. This will comprise a methodological assessment of the impact that changes to the sample size and/or other aspects of the sample design will have on the survey results. It will also involve a practical consideration of options to boost the sample size where appropriate and feasible (eg. potential changes to the frequency).
- Investigations to identify costings and data quality ramifications associated with producing some SMVU data on a biennial basis (issue 2.2)
- Investigate the feasibility of increasing the sample size of non-overlapping samples to improve the accuracy of level estimates and therefore the accuracy of the associated movement between years. Also consider the option of introducing a level of overlap in the sample. (issue 2.3)
- ABS to investigate options for improving data accuracy at the level of sub-state dissaggregation important to users (ie. metropolitan and rest of state focus). This will involve considering stratification options and sample size options (also provider load and ABS survey processing costs) as well as the outcomes from the investigation into data frequency. (issue 2.4)
- Investigate the trade-off relating to increased data accuracy with reduced data availability. (issue 2.5)
3. Progress various options for addressing the disparity between SMVU fuel use data and fuel sales data, subject to the results of investigation 1.
- Investigate the potential of producing accurate vehicle type/age delineation of total fuel sales without using SMVU data. (issue 2.6)
- Develop a new strategy/enhanced questioning, with the aim of reducing non-sampling error when producing SMVU fuel use estimates. (issue 2.6)
4. Investigate options for reducing selected aspects of the non-sampling error of the survey estimates.
- Develop a strategy to improve the quality/consistency of the survey frame information supplied by the individual State and Territory Motor Vehicle Registries (MVRs). This will be a long-term project. (issue 2.2)
- Investigate the effectiveness of survey questions known to be problematic for users to answer, and identify possible enhancements where appropriate. (issue 2.9)
5. Undertake a separate analysis to identify the characteristics of unregistered vehicle use (using data from police records, State and Territory MVR authorities, mining and agricultural industry groups) (issue 2.7)
6. If no changes to the survey frequency are implemented, progress techniques for developing and publishing 'smoothed' SMVU data series. (issue 2.2)
7. Investigate aligning the ABS and AustRoads vehicle classifications more closely. (issue 2.8)
8. Investigate the feasibility of including selected additional data items in the SMVU. Start by clarifying the accuracy and frequency requirements. (issue 2.1)
9. ABS to investigate publishing a technical paper, discussing the time series aspects of the SMVU and summarising the changes that have occurred over time, particularly from the period 1995 onwards. (issue 2.2)
Table 1.6.1 Link between the SMVU issues and the recommended investigations to address these issues
|Issue||Recommended investigations to address issue|
|2.1 Authoritative data set||
- Consider options for publishing some data confrontation analyses where the SMVU data is compared over time and with other sources. (investigation 1)
- Joint ABS/BTRE analytical project investigating disparities between SMVU data and industry measures. This will involve analysing SMVU unit record data. (investigation 1)
- Investigate the feasibility of including selected additional data items in the SMVU. Start by clarifying the accuracy and frequency requirements. (investigation 8)
|2.2 Data volatility||
- Conduct investigations to identify costings and data quality ramifications associated with producing some SMVU data on a biennial basis (investigation 2)
- Develop a strategy to improve the quality/consistency of the survey frame information coming from the individual State and Territory MVRs. This will be a long-term project. (investigation 4)
- ABS to investigate publishing a technical paper, discussing the time series and summarising the changes that have occurred over time, particularly from the period 1995 onwards. (investigation 9)
- If no changes to the survey frequency are implemented, progress techniques for developing and publishing 'smoothed' SMVU data series. (investigation 6)
|2.3 Measuring level movements in key variables||
- Investigate feasibility of increasing the sample size of non-overlapping samples to improve the accuracy of level estimates and therefore the accuracy of the associated movement between years. Also consider the option of introducing a level of overlap in the sample. (investigation 2)
|2.4 Geographic patterns of activity||
- ABS to investigate options for improving data accuracy at the level of sub-state dissaggregation important to users (ie. metropolitan and rest of state focus). This will involve considering stratification options and sample size options (also provider load and ABS survey processing costs) as well as the outcomes from the investigation into data frequency (investigation 2)
|2.5 Data frequency||
- ABS methodology unit will investigate the trade-off relating to increased data item accuracy with reduced data availability. (investigation 2)
|2.6 Quality of fuel use estimates||
- Investigate the underlying reasons behind the disparity between SMVU fuel use estimates and data on total fuel sales, consider the differences between these two measures. (investigation 1)
- Investigate the potential of producing accurate vehicle type/age delineation of total fuel sales without using SMVU data. (investigation 3)
- Develop a new strategy/enhanced questioning, with the aim of reducing non-sampling error when producing SMVU fuel use estimates. (investigation 3)
|2.7 Impact of unregistered vehicles and out of scope vehicles||
- Undertake a separate analysis to identify the characteristics of unregistered vehicle use (using data from police records, State and Territory MVR authorities, mining and agricultural industry groups) (investigation 5)
|2.8 Vehicle classification used by the ABS||
- Investigate aligning the ABS and AustRoads vehicle classifications more closely. (investigation 7)
|2.9 Question changes on SMVU survey forms||
- Investigate the effectiveness of survey questions known to be problematic for users to answer, and identify possible enhancements where appropriate. (investigation 4, investigation 3 re fuel use questions)
The review team has identified three further issues beyond the scope of the present review. These were raised as part of the Transport IDP process and discussed in the Issues Discussion Paper. These issues are:
These issues are described in more detail in section 3 of this report.
- Measuring transport activity by industry. Stakeholders indicated this is not a significant issue for the SMVU. Stakeholders who do require an industry dimension are currently using the published commodity detail as a proxy.
- Inclusion of specific transport related environmental measures as part of the SMVU. Stakeholders indicated that the SMVU is not the appropriate collection for measuring specific transport related environmental impacts. (General data accuracy investigations will seek to improve the quality of existing SMVU data items that are used in environmental models, which partly addresses this issue.)
- Measuring transport industry financial performance. Stakeholders indicated that transport industry financial performance is not a significant issue for the SMVU.