|Page tools: Print Page Print All|
2.72. This section of the paper has reviewed the accuracy of Australia's balance of payments statistics. It has focused on the various causes of error and the varying analyses regularly undertaken by balance of payments statisticians to identify and correct for error. These have included:
2.73. It was noted that a small balancing item is no guarantee of accuracy and the ABS has pursued, and is continuing to pursue, quality improvements that lift the quality of estimation for both credit and debit series, without necessarily impacting significantly on estimated balancing items. This has been particularly so in regard to measuring Australia's international trade in services (where the net impact of the improvements has not been large), and to improvements to be implemented shortly in measuring reinvested earnings (and the offsetting reinvestment of earnings) on a quarterly basis with no impact on the balancing item.
2.74. However, significant initial estimates of the balancing item have provided pointers to balance of payments statisticians in understanding the nature of some significant errors in the measurement of capital transactions, and assisted in developing improved methodologies to prevent the recurrence of such errors. Factors that were possibly contributing to the current estimates of the balancing item were considered in some detail.
2.75. Finally, this section included a subjective assessment by balance of payments statisticians of the quality of initial estimates published for major components of the balance of payments. The assessments were made in regard to the estimated value of transactions, and did not necessarily reflect the accuracy that would apply to estimates of the period-on-period movement that the initial estimates imply. This was largely because revisions often come about from methodological improvements that change the level of an estimated series without significantly affecting the movements within the time series.