|Page tools: Print Page Print All|
Feature Article - Bilateral merchandise trade statistics reconciliation: Australia and United States of America, 1991 to 1994
The most significant of these coverage differences are those due to the difference in treatment of low value trade, and the difference in treatment of repairs. Investigations have indicated that the discrepancy due to the remaining coverage differences would be minimal.
The UN recommends that the value of exports be the 'free on board' (FOB) value at the frontier of the country of export. For imports, the UN recommends that the 'cost insurance and freight' (CIF) value at the frontier of the country of import be used, and that the FOB transaction value be collected when possible, as supplementary information.
Australia values its exports on an FOB transactions value basis, and its imports on a 'Customs value' basis. Customs value is the FOB transactions value adjusted for any transaction where the Australian Customs Service considers the FOB transactions value to be an artificially low price relative to the market value. The Customs value is applied at either the point of containerisation (most cases), the point of shipment or the customs frontier of the exporting country, whichever comes first.
The United States values its exports on a 'free alongside ship' (FAS) basis. The value includes all costs incurred in placing the goods alongside the carrier at the port of exportation. The United States values its imports on a 'Customs value' basis, which is generally equivalent to an FAS transactions value basis. The FOB value exceeds the FAS value by the cost of loading the merchandise onto the carrier. The United States Bureau of the Census advised that in practice there is little difference between values reported on FOB and FAS bases.
c) Country classification
The UN recommends that exports be classified by country of final destination, which is the country where the goods are going to be consumed. Both Australia and the United States adhere to the UN recommendation. However, this can be a difficult task as the exporter is sometimes not in a position to know whether the goods are to be further manufactured or otherwise consumed in the country to which they are consigned, or whether they will be traded with yet another country. When the country of final destination is not known at the time of exportation, the exporter declares the country of last shipment (country of consignment) in place of the country of final destination.
For imports the UN recommends that the country of origin, or production, be used to identify the trading partner supplying the goods. Production is defined to exclude minor operations, such as shelling, bottling or labelling that do not alter the essential nature of the goods. Both Australia and the United States adhere to this recommendation.
It is clear from the above definitions that, in concept at least, exports and imports statistics will only be symmetrical between trading partners when exports are shipped directly from the country of origin to the country of final destination. Discrepancies occur when third countries are involved, as with re-exports of merchandise and goods traded through intermediate countries.
Three cases where country attribution principles may require adjustments to be made are outlined below.
i) Re-exports of merchandise
When goods are exported to Australia and are subsequently re-exported to the United States, these goods will be recorded in Australia's exports to the United States but will not be included in United States' imports from Australia. Consequently, for data reconciliation purposes, the value of Australia's re-exports to the United States will need to be subtracted from Australia's export statistics. The same situation applies in the case of United States' exports and re-exports to Australia.
As an example, suppose Australia imports a commodity from a third country, and performs a minor operation such as bottling or labelling prior to exporting the commodity to the United States. Australia would classify the commodity as an import from the third country, and a re-export to the United States. However, the United States would not record the item as an import from Australia. Rather, the commodity would be recorded in United States' statistics as an import from the third country (the country of origin), with country of consignment Australia. For the purposes of the reconciliation, it would therefore be necessary to make an adjustment for the value of the re-exported item.
ii) Trade through intermediate countries
Trade through intermediate countries, or indirect trade, occurs when goods pass through a country other than the originating or destination country, remaining in that country for some length of time before shipment to the destination. Indirect trade can take two forms, namely retrading and trans-shipments.
Ideally, an adjustment should be made for retrading activity, since it appears in the statistics of the importing country only. However, in Australia's import statistics it is not possible to differentiate between trans-shipment and retrading activity. In both cases Australia's import data will record the United States as the country of origin and the country of consignment as other than the United States.
Re-imports are goods originally exported which are imported to the originating country in either the same condition in which they were exported or after undergoing repair or minor operations which leave them essentially unchanged.
Goods of United States origin that have been re-imported into the United States from Australia are recorded under a special commodity code in United States' statistics as imports from Australia. An adjustment needs to be made for these in the reconciliation. By contrast, goods of Australian origin that have been re-imported into Australia are included in Australia's published import figures as imports with country of origin Australia, and do not impact on the reconciliation.
The issue of timing can impact on the comparability of trade statistics. For example, due to the distance between Australia and the United States, there will be a time lag between a shipment departing from one country and arriving in the other country. A shipment departing by sea from the United States in mid December 1993 would be recorded by United States Customs in December and be included in 1993 statistics for the United States. However, the shipment would be likely to arrive in Australia in early January 1994, and be recorded in Australia's 1994 import statistics. To compensate for this, a shipping adjustment is needed in the reconciliation to cover goods exported at the end of one time period, but imported at the beginning of the next time period.
If there is a time lag between the arrival/departure of goods and their recording in a country's trade statistics, this will also impact on the comparability. For example, Australia's imports are recorded statistically in the calendar month in which the import entries are finalised by Customs, so goods included in a month's statistics may not always have arrived in that month. An adjustment can be made for the value of goods arriving in Australia outside the year in which they were recorded. No adjustment is necessary for Australia's exports or for United States exports and imports as these are recorded statistically according to the month of arrival (imports) or departure (exports).
e) Exchange rate
As analysis of bilateral merchandise trade data is carried out in a common currency, the conversion of data using an average annual exchange rate can contribute to data discrepancies. These discrepancies cannot easily be quantified, except in relation to large value transactions.
f) Domestic territory definition
As merchandise trade statistics are collected by the national Customs agency, it is important to know the definition of each Customs territory. For example, the United States' Customs boundary includes Puerto Rico and the US Virgin Islands. However, Australia treats these as separate countries for the purposes of identifying country of origin or final destination for goods. It is therefore necessary to make an adjustment to account for trade between Australia and both Puerto Rico and the US Virgin Islands.
g) Commodity classification
Both the United States and Australian data used in this analysis are classified under the Harmonised System. For particular transactions there may be some differences in classification between the two sets of data. As these differences do not affect the comparisons of total trade between the two countries, no adjustments are required.
Both Australia and the United States protect confidential data by reclassifying the data from Chapters 1 to 97 of the Harmonised System to a dump code. Confidentiality does not, however, affect data comparability at the total trade level and therefore no adjustments are required.
i) Data errors
Reporting errors, data entry errors, and mistakes made during editing and processing of data are likely to occur with any merchandise trade statistics. However, their impact at the total trade level is believed to be relatively minor. Where errors are identified during a reconciliation, adjustments can be made.
j) Other differences
After conceptual adjustments have been made, discrepancies between the United States and Australian data may remain. Comparison of data at the commodity level can lead to the identification of further adjustments. These may include: goods that crossed one country's customs frontier but not the other's; differences in interpretation of the merchandise trade definition; and omissions from one country's statistics.
Outcomes of the four reconciliation studies undertaken are shown in Tables A and B. These tables show the value of Australia's exports or imports, the value of adjustments made, and the value of the United States' imports or exports. Also shown is the residual discrepancy - this is the discrepancy that remains between the two sets of figures after all conceptual and practical adjustments have been made. All figures are in Australian dollars, converted using average annual exchange rates.
In each of the four years studied, the value of United States recorded merchandise trade was less than the value of Australian recorded merchandise trade - both southbound (trade bound for Australia) and northbound (trade bound for the United States).
In 1991, Australia recorded imports from the United States of $11,897 million while the United States recorded exports to Australia of $10,792 million, a difference of $1,105 million. In the same year, Australia recorded exports of $5,369 million to the United States while the United States recorded imports from Australia totalling $5,242 million, a difference of $127 million.
In 1992, southbound trade increased, with Australia recording imports from the United States of $12,379 million and the United States recording exports to Australia of $12,221 million, a difference of $158 million. In the same year, northbound trade declined slightly. Australia recorded exports of $5,134 million while the United States recorded imports from Australia totalling $5,078 million, a difference of $56 million.
In 1993, southbound trade continued to rise. Australia recorded imports from the United States of $13,187 million while the United States recorded exports to Australia of $12,239 million, a difference of $948 million. By contrast, northbound trade further declined. Australia recorded exports of $5,071 million to the United States while the United States recorded imports from Australia totalling $4,876 million, a difference of $195 million.
Southbound trade rose again in 1994. Australia recorded imports from the United States of $14,839 million while the United States recorded exports to Australia of $13,435 million, a difference of $1,404 million. In the same year, Australia recorded exports of $4,651 million to the United States while the United States recorded imports from Australia totalling $4,398 million, a difference of $253 million.
The initial discrepancy (the difference between each country's statistics before adjustments have been made) has varied over the four years with the southbound discrepancy changing from 9.3 per cent of Australia's imports in 1991, to 1.3 per cent in 1992, 7.2 per cent in 1993 and 9.5 per cent in 1994. The northbound initial discrepancy changed from 2.4 per cent of Australia's exports in 1991, to 1.1 per cent in 1992, 3.8 per cent in 1993 and 5.4 per cent in 1994.
Analysis of Adjustments
As mentioned in the first part of this article, there are many reasons why Australia's statistics and the United States' statistics may not match. This section of the article quantifies these reasons and compares them over the four years.
These adjustments represent the changes needed to transform Australian published imports to the same basis as United States' published exports:
Adjustments were made to account for the difference in treatment of the value of repairs and low value records.
Adjustments for the difference between Australia's import value on an FOB basis and Australia's import value on a Customs value basis were made in three of the four years studied. In 1992, the FOB value was lower than the Customs value by $8 million, in 1993 the difference was $2 million and in 1994 the difference was $7 million.
The following adjustments were applied to adjust for differences in country attribution principles.
The timing adjustment accounts for merchandise which is likely to have been recorded in different years in the statistics of the exporting and importing countries. It is made up of adjustments for either end of the reference period. A negative adjustment indicates that the adjustment for the beginning of the reference period (for goods that were included in United States' statistics in the previous year, but in Australia's statistics in the reference year) is larger than the adjustment for the end of the reference period (for goods that were included in United States' statistics in the reference year, but in Australia's statistics in the following year).
Negative timing adjustments of $423 million and $18 million were made in 1991 and 1992 respectively, while positive adjustments of $179 million and $63 million were made in 1993 and 1994 respectively. Methodological changes, applied for the first time in the 1993 reconciliation, contributed to the differences in the southbound timing adjustments. These changes are described below.
When calculating the 1991 and 1992 southbound timing adjustments, it was assumed that it took, on average, 21 days to ship goods by sea from the United States to Australia. Further, it was assumed that approximately 60 per cent of Australia's January imports from the United States arrived by sea. For the purposes of calculating the 1993 and 1994 shipping adjustments, the average journey time between the United States and Australia was revised upwards from 21 days to 34 days, based on a detailed analysis of shipping schedules. Further, actual sea trade figures, rather than estimates, were used to calculate the shipping adjustment.
A processing adjustment, to account for the value of goods arriving in Australia outside the year in which they were recorded in Australia's trade statistics, was also made for the first time in the 1993 reconciliation, and then again in the 1994 reconciliation. Processing adjustments of $25 million and $23 million were included as part of the southbound timing adjustment in 1993 and 1994 respectively.
If the 1993 timing adjustment had been calculated using the 1992 methodology, the adjustment would have been $142 million compared with the $179 million calculated using the revised methodology.
The exchange rates used in the reconciliation studies were $US = $A1.2842 for 1991, $US = $A1.3768 for 1992, $US = $A1.4789 for 1993 and $US = $A1.3736 for 1994. Discrepancies due to the conversion of data using an average exchange rate can be easily quantified only in relation to large value transactions. In the reconciliation studies, all transactions in aircraft that were included in both Australia's and United States' statistics were analysed. This adjustment represents the valuation difference attributable to the use of average annual exchange rates in these transactions. The adjustments amounted to $6 million, $30 million, $12 million and -$14 million in 1991, 1992, 1993 and 1994 respectively.
This adjustment represents the value of Australia's published imports from Puerto Rico and the US Virgin Islands. This trade is included in United States' exports but not included in Australian imports from the United States. The adjustments made to 1991, 1992, 1993 and 1994 amount to $241 million, $195 million, $225 million and $234 million respectively.
The following additional adjustments were made to account for other differences detected during the reconciliation.
The residual discrepancy shown in Table A represents the difference remaining after the application of the adjustments described above. The discrepancy amounts to $295 million, $87 million, $328 million and $711 million in 1991, 1992, 1993 and 1994 respectively. For each of the four years studied the residual discrepancy is negative, indicating that the adjusted Australian merchandise imports figure is higher than the published United States' merchandise exports figure.
Possible reasons for the residual discrepancy include: imports through third countries of goods of United States origin not identified as such in Australia's imports; valuation differences; additional timing differences; minor coverage differences; currency conversion practices; and non-filing of export documents by some United States exporters.
These adjustments represent the changes needed to transform Australian published exports to the same basis as United States' published imports:
The only adjustment made is for the difference in treatment of low value records, as the value of repairs was minimal.
This adjustment represents the difference between the United States' estimate for trade below $US1,251 and the actual value of Australia's export transactions between $A500 and $US1,251. The effect of this is to remove the value of all transactions less than $US1,251 from the reconciliation. This adjustment was made for the first time in 1993 and the adjustments made were $6 million and $10 million in 1993 and 1994 respectively.
The following adjustments were applied to adjust for differences in country attribution principles.
This adjustment was calculated by the United States Bureau of the Census, and represents United States' imports in 1994 that were exported from Australia in 1993, less United States' imports in 1995 that were exported from Australia in 1994. Positive timing adjustments of $97 million, $80 million and $46 million were made in 1991, 1992 and 1993 respectively. In 1994 a negative timing adjustment of $56 million was made.
This adjustment represents the value of Australia's published exports to Puerto Rico and the US Virgin Islands. The United States includes Puerto Rico and the US Virgin Islands as part of its domestic territory for the purposes of trade statistics, but Australia treats them as separate countries. Adjustments of $6 million, $6 million, $4 million and $6 million were made in 1991, 1992, 1993 and 1994 respectively.
The residual discrepancy shown in Table B represents the difference remaining after the application of the adjustments described above. The discrepancy amounts to $226 million, $115 million, $200 million and $294 million in 1991, 1992, 1993 and 1994 respectively. For each of the four years studied the residual discrepancy is negative, indicating that the adjusted Australian exports is higher than the published United States' imports.
Problems associated with correct country attribution are a likely explanation for at least part of the discrepancy. For example, United States importers may not be able to correctly identify the country of consignment for either re-imports of United States origin goods, or imports of Australian origin goods from third countries. Similar difficulties may occur with regard to trans-shipment of goods of Australian origin, particularly via Canada. In addition, Australian exporters may not be able to correctly determine whether goods being exported are Australian produce or re-exports.
Other possible reasons for the residual discrepancy are: valuation differences; additional timing differences; minor coverage differences; and currency conversion practices.
This series of reconciliation studies has demonstrated that a significant part of the 'asymmetry' in United States-Australia bilateral merchandise trade data results from the conceptual factors underlying the compilation of the data. Actual data errors in the compilation process proved to be relatively insignificant in comparison. As previously indicated, the adjustments presented in the reconciliation do not represent revisions to the official published statistics of either country, nor do they imply, in general, errors in either country's published statistics.
For southbound trade the residual discrepancy is less than the initial discrepancy in each of the four years. In 1991 the initial discrepancy was 9.3 per cent of Australia's imports, and the residual discrepancy 2.5 per cent. The discrepancy in 1992 was reduced from 1.3 per cent to 0.7 per cent, in 1993 from 7.2 per cent to 2.5 per cent and in 1994 from 9.5 per cent to 4.8 per cent. The largest contributors to the narrowing of the gap were the adjustments for Australia's indirect imports of United States origin goods and the underestimation of low value trade by the United States.
For northbound trade the residual discrepancy is greater than the initial discrepancy in all four reconciliation studies. In 1991 the discrepancy rose significantly from 2.4 per cent to 4.2 per cent after application of the adjustments. In 1992 it rose from 1.1 per cent to 2.2 per cent. In 1993 and 1994 the northbound discrepancies showed less change, rising from initial discrepancies of 3.8 per cent and 5.4 per cent to discrepancies after adjustment of 3.9 per cent and 6.3 per cent respectively. The most significant factors contributing to the widening of the gap on northbound trade were the adjustments for United States' indirect imports of Australian origin goods and United States' re-imports. Over the four years studied, both indirect imports and re-imports rose, even though United States' total imports declined.
Further reconciliation studies with the United States are not planned at this stage. However, work has commenced on reconciliation studies with New Zealand, Japan, Indonesia and Korea. Results of these studies will be published in later issues of International Merchandise Trade, Australia (ABS Catalogue Number 5422.0).
These documents will be presented in a new window.