|Page tools: Print Page Print All|
2.62. Taking into account all the factors mentioned so far, it is possible to form some subjective assessment of the quality of balance of payments statistics. Assessments based on recent experience with current data sources and methodologies are provided in Table 5. To give an idea of the relative importance of each item, 1993-94 values are also provided. The assessments relate to the accuracy of the first-published monthly, quarterly and annual estimates of the principal balance of payments aggregates. The initial monthly and quarterly estimates appear in the monthly (Cat. no. 5301.0) and quarterly (Cat. no. 5302.0) publications, respectively. Initial annual estimates are a little different. The first estimates of current account items and some limited capital account items are released in the June monthly publication each year, while the remaining capital account items appear first in the June quarterly publication each year.
2.64. Those given a B rating (the estimate is considered to lie within 10% of the true value) are:
2.65. Other estimates are considered less accurate and have been rated as C (within 15% margin of error) or D (15% or greater margin of error). The aggregates (not balances) considered to be least accurately estimated are:
2.66. It should be stressed that the ratings used are assessments of the quality of the transactions estimates published in the accounts and do not necessarily reflect the quality of estimated movements in the various items and aggregates shown. For example, the initial monthly estimate of the current account deficit is shown in Table 5 with a rating of D (an error margin of 15% or more). However, analysis later in this paper (under the heading Do revisions alter information provided by initial estimates?), concludes that the month-to-month movements in the current account deficit present a relatively reliable series without obvious bias.
2.68. Problems with both the coverage and measurement of international capital flows of the non-official sector are a particular concern, generally reflecting the deregulation of capital markets since the mid-1980s. The earlier discussion under the heading Contributions to the balancing item outlined some of the specific problems that exist and some of the actions being taken by the ABS to address these problems.
2.69. Problems also exist with the measurement of some services, income and unrequited transfers items, but these are considered to be on a much smaller scale than those in the capital account. An example is the international trade in other services (other than transportation and travel services) which grew strongly during the 1980s. For a number of reasons this is a difficult area about which to compile statistics, particularly on the debits side. The types of services involved are diverse while the service providers and receivers are widespread throughout the economy and continually changing. There is no comprehensive source of coverage of either transactors or transactions. There are also conceptual and practical difficulties involved in identifying internationally-traded services without omission or overlap with other balance of payments items. Nevertheless, the ABS has made considerable strides in improving data in this area in recent years. An article on Developments in Measuring Australia's International Trade in Services, published in International Trade in Services, Australia, 1992-93 (5354.0) describes the data sources currently used, changes made in the collection strategy, data quality initiatives being pursued and the future developments that are expected.
2.70. Problems currently exist with the estimation of the transfers made by non-residents migrating to Australia (included in unrequited transfers credits). In estimating this item, reliance is placed on various assumptions about the value of funds and other assets transferred and whatever data are at hand (such as occasional surveys). A longitudinal survey of immigrants being conducted by the Bureau of Immigration, Multicultural and Population Research is expected to provide the basis for much more soundly-based estimates. Preliminary results from this survey are currently being evaluated.
2.71. The accuracy of investment income estimates in the current account may also be influenced by difficulties in the measurement of related capital account items. To the extent that there has been undercoverage of financial claims, it is likely that there will also have been understatement of the associated income transactions.