Annex 9.A. Statistical data validation system of Brazil (example for exports)
9.A.1. The statistical validation system for Brazil’s exports is applied in three different contexts: (a) at date of entry (through dialogue) before the export register (declaration) has been actually produced (preventive validation) (b) continuously, during the operations ( involving export registers with significant values and other criteria) and (c) posteriori (post-validation), after the export operations have been completed.
1. Preventive validation
9.A.2. The ultimate goal of the system is to validate specific fields of the export record that are of statistical concern, in order to achieve the highest possible level of accuracy inBrazil’s statistics of foreign trade. To reach this goal, the system sets parameters for FOB-type unit values (both in kilograms and in supplementary quantity units) of the Southern Common Market (MERCOSUR) Common Nomenclature (NCM) codes. These parameters are set using data for at least six months and a statistical model that: (a) takes into account the number of observations, (b) eliminates peak and bottom distortions, (c) establishes middle values (by kilogram and supplementary quantity unit), (d) establishes the standard deviation and maximum and minimum quantity thresholds (by kilogram and supplementary quantity unit) and (e) creates a table of parameters by commodity (NCM and HS code).
9.A.3. The FOB-type value inUnited Statesdollars, the net weight in kilograms and the supplementary quantity are used to calculate the average value. In order to avert a slowing down of operations or the creation of difficulties for the exporter, there is a tolerance for minimum and maximum values of one decimal (-90 per cent and +900 per cent). If the operation is within these margins, the exporter is informed that there may be an error, but the operation is not stopped. Outside these margins, however, the export record is forwarded (centralized) for analysis to the appropriate department in the Secretariat of Foreign Trade (SECEX). The Secretariat can communicate through the system with the exporter, who then needs to ratify or justify the data presented in his Export Register.
9.A.4. When the exporter enters data intoBrazil’s Integrated System of Foreign Trade (SISCOMEX), the system analyses and compares them to the middle value by commodity. This preventive validation stage leads to various possible scenarios, as illustrated in box A9.1.
9.A.5. Most of the reasons for forwarding (centralizing) an export record for statistical analysis involve errors in the following information: (a) FOB value in United States dollars, (b) net weight (in kilograms), (c) supplementary quantity, (d) commodity code (NCM–HS) or, (e) currency unit or, (f) the INCOTERMS rule applied. The centralized review of export records constitutes approximately 1 per cent of 20,000 daily operations for exports. More than 99 per cent of these centralized records have typing errors.
2. Continuous and post-validation
9.A.6. Continuous validation is executed using management reports, which cover the main daily operations. It is based on the “significant value principle” and the use of comparative statistics.
9.A.7. Post-validation is executed after the conclusion of the export operation (shipment), based on validation reports entailing, e.g.:
(a) Comparison between net weight and value;
(b) Checking of the relation between merchandise by country for past periods;
(c) Analysis of the export registers flagged by yellow and red occurrences.