COHERENCE
11.50 Coherence refers to the internal consistency of data at a point in time and over time, as well as its comparability with other sources of information.
11.51 Coherence is an important component of quality as it provides information about any limitations with comparing the data with other data sources. The use of standard concepts and definitions and robust methods promote coherence. Coherence with other statistics does not on its own mean the statistics are accurate, as the statistics that are being compared may suffer from similar sized errors in the same direction.
CONSISTENCY CHECKS AND PROCESSES
11.52 International merchandise trade statistics are almost exclusively compiled from customs declarations. This together with a robust processing system (with separate environments for testing and training) and established processes for receiving, editing, aggregating, confidentialising, finalising and disseminating data ensures a high level of coherence.
11.53 There are a range of checks applied at various stages in the ABS compilation processes to ensure the consistency of the data. For example, there are edits which may identify errors in either the gross weight or quantity when the unit of quantity is grams, kilograms or tonnes. Other edits applied at the individual transaction level check the validity of the relationship between related fields, for example, if the mode of transport is sea but the goods were loaded in Switzerland. These inconsistencies are detected and corrected prior to the release of initial estimates.
11.54 International merchandise exports and imports are one of the first releases of monthly economic statistics so there are limited published sources available to confront against the estimates. Consequently, considerable attention is given to ensuring the consistency of the monthly aggregates at the HS 6 digit and SITC 2 digit level with previous monthly results. Unusual movements may be verified or corrected using publicly available information (e.g. press reports, company websites) or following contact with the exporter, importer or their agent.
11.55 Because the ABS publishes a large range of economic data it strives to ensure coherence across these estimates. To achieve this processes have been implemented including regular quarterly meetings between representatives from the various economic collections. At these meetings there are discussions which lead to agreement on the statistical treatment (including valuation, timing and classification) of recent economic events such as major resource projects.
11.56 Any significant changes to the established processes must be reviewed and approved by a Methods Board. The Methods Board consists of a panel of ABS subject-matter experts who are responsible for endorsing new methodologies, and enhancing the coherence of statistics across the ABS.
INTERNATIONAL COMPARABILITY
11.57 The HS was adopted internationally by the Customs and Co-operation Council in 1983 and entered into force in January 1988. The DIBP and the ABS have used the HS classification since that time so comparison of the data (up to the 6 digit level) with other countries using the HS is possible from that date. The ABS also publishes international merchandise trade statistics by SITC for comparison at the international level. Confidentiality restrictions can change the availability of data at any point so care should be taken when comparing data over time and with other countries. Confidentiality restrictions are not applied to the value of total trade (exports, imports or import clearances).
11.58 Australia's international merchandise trade statistics are compiled substantially in accordance with IMTS 2010 which is also broadly consistent with 2008 SNA and BPM6. Australia, like many of its trading partners, publishes the details of its compliance with the international standards so statistical users are able to make informed decisions about the level of data comparability. Australia's proposed stance against the international standards are set out in the Information Paper: Proposed Implementation of the New International Standard for International Merchandise Trade Statistics (cat. no. 5368.0.55.020). Reasons for differences in the statistics released by each country are also available, for example, IMTS 2010 recommends that imports are valued on a CIF basis. While Australia does compile imports on a CIF basis and have these available for release, the published imports valuation is customs value (a FOB type valuation, see paragraph 3.12 in Trade System) which is the value on which customs duty is applied and is of a higher quality than the CIF value at the commodity level. In Australia, a FOB type value is used for both international merchandise trade, balance of payments and national accounts statistics so coherence between these data is maintained. However, comparisons of Australia's exports to a country which publishes imports (from Australia) on a CIF basis will be affected because of the different valuation basis (as well as other differences including timing, coverage (e.g. lower or higher thresholds) and country classification).
11.59 The ABS has conducted bilateral reconciliations of international merchandise trade data with the United States, Japan, New Zealand and the European Union. These studies demonstrated that the main reasons for differences in the statistics of Australia and its major trading partners are due to the conceptual and methodological factors underlying the compilation of the data. The ABS does not have the resources to conduct further bilateral reconciliation studies. Statistical users interested in conducting their own bilateral studies should see:
Print Page
Print All