EDICOM REPORT. Analysis of Asymmetries in intra-community trade statistics with particular regard to the impact of the Rotterdam and Antwerp effects

Transcription

1 EDICOM REPORT Analysis of Asymmetries in intra-community trade statistics with particular regard to the impact of the Rotterdam and Antwerp effects December 2005 Statistics and Analysis of Trade Unit HM Revenue & Customs 3rd Floor Alexander House 21 Victoria Avenue Southend-on-Sea SS99 1AA United Kingdom

2 Table of Contents 1. Introduction The equipment and software applications The Rotterdam Effect Research into the Rotterdam Effect via another Country via another Country Asymmetries EU Regulation Changes Recommendations Conclusion Annex A Annex B Annex C References

3 1. Introduction The aim of this study was to analyse the asymmetries in intra-community trade statistics with particular regard to the impact of the Rotterdam/Antwerp effect, and to provide recommendations on improving EU trade statistics and better align them with international definitions. Asymmetries between the, and and the were identified at the chapter level for 2002 and 2003, and at the aggregate monthly level for 2000 to Research was also undertaken on the Rotterdam Effect; the different kinds of transit flow and what impact these have on trade statistics. This study was undertaken by the Statistics and Analysis of Trade Unit (SATU). SATU is part of the s HM Revenue & Customs (HMRC), and is responsible for the collection, compilation and publication of the s Overseas Trade Statistics (OTS). 2. The equipment and software applications Data were taken from the Comext Database via Internet link and are on a Special trade basis. The data were manipulated in SAS and Excel. 3

4 3. The Rotterdam Effect The UN International Merchandise Trade Statistics Concepts and Definitions (IMTS) (United Nations, 1998 paragraph 45) defines goods in transit to be goods entering and leaving a country with the exclusive purpose of reaching a third country. The Rotterdam Effect, alternatively known as the Antwerp Effect, is a term used to describe asymmetries and other anomalies arising from transit trade through major ports, such as Rotterdam, whether these arise from traders correct or incorrect declarations. Principally, the Rotterdam Effect causes imports and exports to be attributed to the country of transit as opposed to the real partner country. The Rotterdam Effect is not confined to trade between EU Member States but can affect trade between any pair of countries where the goods are transported through one or more additional countries. Transit Flows There are several different kinds of transit flow, and it is how these flows are collected and recorded in trade statistics that may lead to the Rotterdam Effect. Non-EU Transit Trade For exports from the EU that transit through a non-eu country, figure 1 shows the possible forms this transit might take. Figure 1: via non-eu EU Country of Dispatch Non-EU Country Of Destination Non EU Transit Country Customs Border Simple Transit Indirect Trade Re- Free Circulation Property of Resident 4

5 Figure 2 shows the cases for imports from outside the EU that pass through another non-eu country. Figure 2: via non-eu EU Country of Destination Non-EU Country Of Despatch Non EU Transit Country Customs Border Simple Transit Indirect Trade Re- Free Circulation Property of Resident In these diagrams three main kinds of transit trade are identified: simple transit, where the goods pass through a country and not declared as imports or exports in the country of transit; indirect trade, where goods are cleared as imports and then subsequently exported again, but always in foreign ownership; re-exports, similar to indirect trade, but the goods become the property of a resident in the state of transit on import, and then resold on export. 5

6 Transit Trade within the EU Figures 3 and 4 show the differences involved where the country of transit is within the EU. Note it now makes a difference whether the partner country at the end of the chain is also within the EU or not. Figure 3: via the EU EU Country of Despatch Free Circulation Customs Border Customs Border EU Transit Country Simple Transit Indirect Trade Re- Non-EU Country Of Destination EU Country of Destination Property of Free Resident Circulation Simple Transit Figure 4: via the EU EU Country of Destination Free Circulation Customs Border Customs Border EU Transit Country Simple Transit Indirect Trade Re- Non-EU Country Of Despatch EU Country of Despatch Property of Free Resident Circulation Simple Transit 6

7 Again, the same three kinds of transit can occur with an EU country of transit: simple; indirect; and re-exports. However, since goods can move freely between EU countries without customs formalities, i.e. directly into free circulation, there are differences with movements across the customs border. Most notably, simple transit between EU Countries takes place entirely on the free circulation side of the customs border. 7

8 4. Research into the Rotterdam Effect At the 6 th OECD International Trade Statistics Expert Meeting in Paris Mr. Jasper Roos of Statistics gave a presentation on International Transport and Trade Statistics. As covered in the accompanying paper (Roos, 2005), he identified five kinds of transit flow through the : re- the export of previously imported goods, which became the property of a Dutch resident on import; quasi-transit the export of previously imported goods, which were never the property of a Dutch resident; transit through a customs warehouse; transit through the goods that pass through the but are never stored, traded or cleared in the ; and transito goods bought and then sold by a Dutch resident, but the goods themselves never enter the. Quasi-transit is clearly the same as indirect trade. Roos identifies quasi-transit as the Rotterdam effect. Transito is commonly known as triangular trade. According to Roos, only the re-export category really belongs in international trade statistics, but quasi-transit is also included by the EU. However, he does admit that the conceptual differences between re-export and quasi-transit is in practice quite narrow, especially when this involves multinational corporations. Based solely on the compulsory fields required by Eurostat for both Intrastat and Extrastat, these flows cannot be distinguished. The is able to distinguish between re-export, quasi-transit and normal exports for their exports since they require an additional Statistical Procedure field to be filled in (Statistics, 2005), which does effectively encode re-exports and quasi-transit. For the, on a yearly basis, quasi-transit trade amounts to about 30 billion (Roos, 2005). Impact on Trade Statistics Since the Rotterdam effect is based around the misattribution of partner country data, it is a cause of asymmetries between countries trade data. Therefore, several bilateral asymmetry studies between the EU and non-eu countries have identified the Rotterdam Effect as a component of the asymmetries. In 2002, the Australian Bureau of Statistics and Eurostat carried out a bilateral asymmetry study (International Merchandise Trade, Australia, 1998). This notes that EU imports from outside the EU are attributed to the Member State where the goods are entered into free circulation. This is often the port of first entry into the EU, whereas the Australian export declaration declares the final EU country of destination. The asymmetries of Australian exports to the imports of the individual Member States were significantly greater than with the EU as a whole. This is described as the Rotterdam Effect. However, the report also describes other difficulties with Country Classification. There are Australian re-exports, where Australia declares these as exports to the EU, but in the EU records these as imports 8

9 from the country of origin. Similarly, there are EU indirect imports, where goods of Australian Origin are imported and then re-exported via another country, which are declared in the EU as imports from Australia but are only recorded as Australian exports to the EU if the exporter fills in an EU country of final destination. Aside from asymmetries, when Trade Statistics are published and analysed, the Rotterdam Effect is sometimes referenced to explain unusual properties of the data. For example, in January 2003 an article on China s trade with the EU (Bautier, 2003), notes that Dutch imports from China are overestimated due to the Rotterdam Effect, which also has the effect of improving the external trade balances of those Member States which are the true destination of the goods. Similarly, the United Nations Economic Commission for Europe (UNECE) Timber Committee s Australian Market Report, 2002 (UNECE, 2002) described how Austria s timber imports were boosted as much by the Rotterdam Effect, whereby the imports were destined for other EU Member States, as by actual growth in the internal market. However, it is not only where the country of transit is an EU Member State that the Rotterdam Effect can come into play and cause asymmetries. In 2000, Statistics Canada and Eurostat carried out a bilateral asymmetry study (International Trade Division Statistics, Canada, 2000). This study draws attention to the difference between indirect trade and transit trade. By their definitions, in the former, the goods are imported into an intervening country and then re-exported, as opposed to the latter whereby the goods merely move in bond through the intervening country. As such, for indirect trade the final destination may well not be known. This indirect trade is equivalent to the quasi-transit trade identified by Roos. In terms of the impact on Canadian/EU asymmetries, the main intervening country is the USA. Thus for Eastbound trade, the Canadians class the indirect trade as exports to the USA, and the transit trade as exports to the EU. The EU classes both of these as imports from Canada on a country of origin basis. Similarly for Westbound trade, the EU may record exports to the USA whereas Canada records an import from the EU due to country of origin. In 2004, HMRC studied asymmetries with the USA and Canada (Terrazzano, 2004). Here it was discovered that crude oil is imported from the via Portland, Maine in the USA, and then sent by pipeline to Canada. This is recorded in Canada as in import from the, whereas the records an export to the USA. The USA do not record either flow, as it is simple transit. Thus the records more oil exports to the USA then they record importing, and recorded less oil exports to Canada then they record importing. The Introduction of Intrastat Since some of the issues involved with the Rotterdam effect are connected to the interface between the Intrastat and Extrastat trade data collection systems, it is useful to look at how the introduction of Intrastat in 1993 influenced the Rotterdam Effect. Fortunately, in August 1994 the report Visible Trade Statistics The Intrastat System was published in Economic Trends (Williamson and Porter, 1994). Here Rotterdam Effect is used to refer to an anomaly from before the introduction of Intrastat. Prior to Intrastat some imports from non-eu countries which came via other 9

10 Member States before clearing customs in the were incorrectly recorded as imports from that other Member State. After 1993 and the introduction of Intrastat these movements were correctly recorded since it was known that Member States should not appear on Extrastat declarations. Williamson and Porter estimate that about 2 billion of imports in 1992 were erroneously declared as coming from within the EU, of which about 0.8 billion was from the. A related effect in the paper is referred to as hidden transit, which are imports cleared for free circulation in another Member State before arriving within the. This is essentially the same as the indirect trade referred to in the previous two papers. It was feared that these movements might not be captured by Intrastat due to traders not understanding their obligations. However, this was found not to be the case. Implications of Research So the Rotterdam Effect comes into play with transit both within and without the EU. The tendency is for an export to be declared to the goods final destination, whereas the import is declared in that final destination as coming from the transit country. But what does legislation require and international recommendations suggest about the correct declaration of these movements? 10

11 5. via another Country The IMTS (United Nations, 1998) provides guidance and best practice on the compilation of trade statistics, while EC Legislation governs the specific compilation of trade statistics within the EU. In general, the legislation follows the concepts contained in the IMTS. For transit flows, we shall look at what the IMTS guidance for the declaration of transit flows, and then see how the implementation of EU Legislation differs from this. IMTS Recommendations The IMTS (United Nations, 1998) recommends that, for an export, the country of last known destination be recorded as the partner country (paragraph 150). Paragraph 45, Goods in transit, says Goods entering and leaving a country with the exclusive purpose of reaching a third country are excluded. For an export or dispatch from a Member State to another country via an intermediary country, whether simple transit or indirect trade, the recommendations say that the final destination should be recorded, irregardless of whether the two other countries involved are themselves Member States or not. EC Legislation Where the final destination is outside the EU, article 161 (5) of Regulation (EC) 2913/92 clearly states that The export declaration must be lodged at the customs office responsible for supervising the place where the exporter is established or where the goods are packed or loaded for export shipment. HMRC s Notice 275, Procedures, also reinforces this point, saying EC law requires that, normally, the export declaration must be lodged in the exporter s Member State. Only in exceptional circumstances may the declaration be lodged in another Member State. This is in line with the IMTS recommendations. Also, once an export declaration has been lodged, no Intrastat declaration liability arises for movement through another Member State, also in line with IMTS. Where the final destination is another Member State, EU regulations simply require the final destination to be recorded, and again no liability for declaration in the EU state of transit. As part of the s contribution to Eurostat s Guidelines on the Implementation of the Intrastat Legislation HMRC has written guidance on the correct procedure for declaring exports (see Annex A). 11

12 6. via another Country IMTS Recommendations The IMTS (United Nations, 1998) recommends in paragraph 150 that, for imports, the partner country be recorded as the country of origin. In paragraph 45, it is recommended that the movement not be recorded in the intermediary country. EC Legislation If the intermediary country is outside the EU, then EC regulations are inline with the IMTS recommendations since the country of origin is recorded on the Customs SAD (Single Administrative Document) declaration. Unfortunately, for where the intermediary country is another MS, there is no equivalent to article 161 for imports. The EU country where the goods are cleared for free circulation will be deemed to have imported the goods from the non-eu country, and any onward movement will be declared under Intrastat. The importer in practice has two choices. They can move the goods from the other Member State to the under Community Transit, and then clear for free circulation in the, or they can clear them for free circulation in the country where the goods first enter the EU. In the first case, the declaration will be in line with the IMTS recommendations. Nothing will be recorded in the other Member State, and the will record an import from the non-eu country. In the second case however, the other Member State will record an import from the non-eu country, and correspondingly an Intrastat Dispatch to the. The will record an arrival from that Member State. So although both the and the other Member State s balance of payments remain unaffected, and asymmetries between the two are avoided, this is not inline with the IMTS recommendations. Compared to those recommendations, the would tend to overstate its arrivals from within the EU with a corresponding understatement of the s non-eu imports. The other Member State would also have an artificial increase in its imports and dispatches. 12

13 7. Asymmetries Comparison with the EU In principle, any asymmetries due to transit trade through another country will not cause a net change in a countries import or export figures, and the asymmetry will disappear when countries are aggregated together. For the EU member states, we will look at the asymmetries before and after aggregating trade to and from the. All asymmetries in this report are defined as reported imports subtracted from reported exports. Table 1: Import Asymmetries with the EU, Total of Individual Absolute Asymmetries Billion Aggregate Absolute Difference Source: Comext Database For Arrivals there is little difference between the two figures in table 1. Notably in 2000, for all EU Member States the arrival figures were lower then the other states dispatch figures, leading to no difference between the asymmetry figures above. This implies that either transit trade is having little distortive effect on the s arrivals from the EU, or that high levels of other asymmetry issues are masking the effect. The table does not include the s estimates of the impact of Missing Trade Intra Community (MTIC) fraud estimates, since these estimates cannot be allocated by country. However, the estimate was 4.1 billion in 2000, 10.4 billion in 2001, 16.5 billion in 2002 and 6.3 billion in So movement in these estimates can partly explain the increase in asymmetries between 2000 and 2001, and the fall between 2002 and

14 Table 2: Asymmetries with the EU, Total of Individual Absolute Asymmetries Billion Aggregate Absolute Difference Source: Comext Database In table 2 quite a marked difference can be seen, implying that transit trade could well be distorting the true destination of goods bound for Europe. The notable jump from only 0.6bn to 6.8bn between 2002 and 2003 is due to a marked drop in the s declared dispatches to Germany. In 2002, the s dispatches to Germany exceeded Germany s arrivals figures by 0.9bn. In 2003, Germany s arrivals figures exceeded the s dispatches by 2.4bn. Table 3: Import Chapter Asymmetries with the EU, HS2 Chapter Total of Individual Absolute Asymmetries Aggregate Absolute Billion Difference Pharmaceutical products Machinery and mechanical appliances Electrical machinery and equipment Vehicles; other than railway or tramway rolling stock Pharmaceutical products Machinery and mechanical appliances Electrical machinery and equipment Vehicles; other than railway or tramway rolling stock Source: Comext Database Looking at the imports of individual chapters, it can be seen that several chapters show noticeably reduced asymmetries with the EU as a whole for the individual countries. In particular for both pharmaceuticals and vehicles aggregating the countries reduces the asymmetry by other two thirds (table 3). MTIC trade involves commodities in chapter 85. The estimate of the impact of MTIC fell by about 10 billion between 2002 and 2003, and so may account for part of the fall in the chapter 85 asymmetries between 2002 and

15 Table 4: Export Chapter Asymmetries with the EU, HS2 Chapter Total of Individual Absolute Asymmetries Aggregate Absolute Billion Difference Mineral fuels and oils Pharmaceutical products Machinery and mechanical appliances Aircraft and spacecraft Mineral fuels and oils Pharmaceutical products Machinery and mechanical appliances Aircraft and spacecraft Source: Comext Database The effect is also quite noticeable for the s exports, with fuel and aircraft showing the most improvement from aggregation. Transportation via pipelines causes particular problems with reporting the correct partner country with chapter 27 (table 4). 15

16 and the For the whole period, the s import asymmetries with both and the have their declared exports to the as being higher then the s declared imports (Figure 5). There is no overall trend in the asymmetries over the four years, however changes in the pattern of MTIC fraud over the period may be masking other effects. If transit trade was a dominating cause of the asymmetries it would be expected that the s declared imports would be higher. These figures overall would indicate that transit trade is not the biggest contributing factor to the s asymmetries with or the. (Full data on the s monthly asymmetries with and the are in Annex B.) 600 Figure 5: Import Assymetries with and the, Million Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct

17 For the s export asymmetries with, the s declared exports have generally been lower then s declared imports, with no overall trend, and a relatively low level of asymmetries. However with the, although for most of the period the imports have been lower then the s exports, there is a clear downward trend, and in December 2003 the imports exceeded the s exports (Figure 6). It is possible that transit trade issues caused the high levels of asymmetry in (Full data on the s monthly asymmetries with and the are in Annex B.) Figure 6: Export Assymetries with and the, Million Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct

18 The major asymmetry chapters for imports are the same for as for the EU as a whole (Table 5). For all of them, s exports are higher then the imports, in line with the overall asymmetry noted earlier. Full details are in Annex C. Table 5: Import Chapter Asymmetries with, HS2 Chapter Billion Pharmaceutical products Machinery and mechanical appliances Electrical machinery and equipment Vehicles; other than railway or tramway rolling stock Pharmaceutical products Machinery and mechanical appliances Electrical machinery and equipment Vehicles; other than railway or tramway rolling stock Source: Comext Database Similar chapters cause problems with imports from the, with the notable exception of chapter 27 (Table 6). Also only for chapter 85, electrical machinery, do the imports exceed the exports. (Full details are in Annex C.) Table 6: Import Chapter Asymmetries with the, HS2 Chapter Billion Mineral fuels and oils Pharmaceutical products Machinery and mechanical appliances Electrical machinery and equipment Mineral fuels and oils Pharmaceutical products Machinery and mechanical appliances Electrical machinery and equipment Source: Comext Database 18

19 Due to the low level of export asymmetries with exports to at the aggregate noted earlier, the major chapters similarly show a fairly low level of asymmetries. Chapters 71 and 85 show large asymmetries in 2002 but not 2003 (Table 7). (Full details are in Annex C.) Table 7: Export Chapter Asymmetries with, HS2 Chapter Billion Mineral fuels and oils Precious Metals, stones and jewellery Machinery and mechanical appliances Electrical machinery and equipment Mineral fuels and oils Precious Metals, stones and jewellery Machinery and mechanical appliances Electrical machinery and equipment Source: Comext Database As with the import asymmetries with the, chapter 27 is a major cause of asymmetries (Table 8). Also, the larger exports compared to the imports indicates a possible transit trade issue. (Full details are in Annex C.) Table 8: Export Chapter Asymmetries with the, HS2 Chapter Billion Mineral fuels and oils Books, Newspapers etc Machinery and mechanical appliances Electrical machinery and equipment Mineral fuels and oils Books, Newspapers etc Machinery and mechanical appliances Electrical machinery and equipment Source: Comext Database 19

20 Specific movements, 2003 Energy products like oil, gas and electricity can cause particular problems with transit trade, due to often being continuously imported and exported via pipelines or pylons, and not in discrete shipments, so we would expect to see mismatches with these countries. Table 9: Oil Asymmetries with the EU For 2003 Country Import Million Export Austria Germany 375-1,491 Denmark Spain Finland 2 29 France Greece -5-6 Ireland Italy Luxembourg ,297 Portugal Sweden Source: Comext Database For imports compared to other MS s exports, all large asymmetries show the imports to be lower then the MS export, so cannot be used to discern transit trade (Table 9). By comparison, for exports the exports to the exceed their declared imports by over 1 billion. Whereas the s exports to Germany are under Germany s imports by a similar amount. This would indicate transit trade, and that some oil bound for Germany is declared instead as an export to. The has gas pipelines with the, and Ireland. The declared exports of 366 million to Ireland in 2003, but Ireland only declared imports of under half this, 172 million. 20

21 The s pipeline to the goes directly form installations in the North Sea, and so cannot be used for, so the exports of 33 million that the declared to the most likely was transit trade that used the pipeline (table 10). Table 10: Gas Asymmetries with the EU For 2003 Country Import Million Export Ireland Source: Comext Database All electricity imports and exports between the and mainland Europe use a single connection via France. The s electricity declarations for 2003 only have France as a partner country. Nonetheless, several other MSs declare imports and exports of electricity from the, so it would be expected that some of the s declarations involving France are transit trade. However, the s declarations are lower then France s for both imports and exports (Table 11), indicating other problems in the data. Table 11: Electricity Asymmetries with the EU For 2003 Country Import Million Export Austria Germany 1-2 France Greece 0 0 Ireland Source: Comext Database Combining countries, Where the transit flows are known, data from the countries involved can be combined to study the impact on asymmetries. We did this earlier at a macro level by grouping together all the EU countries, and we saw that for the s oil exports the near mirroring of the asymmetries with the and Germany. If the oil trade 21

22 with the and Germany are combined, the asymmetry is reduced to only 193m, compared to a total declared export of 6.4 billion. Other data sources In the, VAT returns from traders include two additional boxes on the value of trade with other member states. As part of the VAT return, which firms complete each quarter, there is a declaration of the total value of exports of goods (Box 9) to customers in other EU Member States and the total value of imports of goods (Box 8) from suppliers in other EU Member States. These returns provide a direct estimate of the size of exports to and imports from the EU. They are also used to identify the requirement to fill in a more detailed, monthly Intrastat return (if the value of the declarations on the VAT return exceed the Intrastat threshold). In addition to the data available from the VAT returns, traders who have arrivals or dispatches with other member states must also submit a European Sales List (ESL). This is a mandatory document that all VAT registered traders are required to complete. The trader is required to supply the details of each dispatch transaction that they have made with a trader in another Member State stating the partner trader's VAT Registration number and the value of the goods (but with no product description). EU legislation sets out the VAT control arrangements for administrative co-operation between Member States. This requires each Member State to maintain an electronic database of data from ESLs and to share and exchange this data with Member States in order to confirm the validity of VAT registration numbers. This exchange scheme is known as VIES (VAT Information Exchange System). It is a condition of zero-rating the exporter's supply that the VAT registration number of his EU customer is correct, so information provided on the ESL is a means for businesses to ensure that they are accounting for VAT correctly and confirms their entitlement to a VAT repayment. Since the completion of an ESL is mandatory, it should be a comprehensive list as it will record dispatches by every VAT registered trader. But the extent to which this requirement is regulated and is enforced is likely to vary considerably across Member States. There is evidence to suggest that in some cases the penalty for non-compliance is lower than the cost of employing someone to complete an ESL, so traders may not be concerned about complying. Information from ESLs is also used to prevent the wrongful diversion of home-use goods that traders have zero-rated as exports. Using information from ESL submitted in other Member States could also help identify potential MTIC fraud (a further cause of asymmetries. Exchanging information and detailed level data with other member states in bilateral projects can identify discrepancies in data, but is very time consuming and cannot be included as part of the checking procedures before the data are first published. However, the has begun to collect additional data on warehousing and transit trade directly from businesses. This enables them to adjust their trade figures to take account of transit trade. 22

23 Tobacco products One area of interest is excise goods, that is goods subject to extra taxes. These goods are subject to additional controls, but also likely to be smuggled or be used in diversion fraud 1. Excise goods are moved via special customs warehouses. While smuggled goods are not included in the trade-in-goods statistics, goods that are diverted may cause problems with asymmetries that add to transit trade patterns. Looking at tobacco goods (HS chapter 24), there is a sudden change in the s asymmetry with Germany, and the (Table 12). This is largely due to a change in the dispatch figures, and not from a decrease in the rest of the EU s arrival figures. The increase in dispatches to and the in 2003 largely corresponds to the fall in despatches to Germany; this may be evidence of diversion fraud. If this was really a change to effectively reporting the country of transit, we would expect the original asymmetry with Germany to be much more neutral. Table 12: Tobacco Asymmetries with the EU, EU Arrivals Dispatches Arrivals EU Dispatches Austria Germany Denmark Spain Finland France Greece Ireland Italy Portugal Sweden Source: Comext Database 1 Goods may either be diverted within the country recording the dispatch or on arrival in the partner country. 23

24 8. EU Regulation Changes In 2005, the EU introduced new regulations regarding trade statistics between Member States, with Regulations (EC) 638/2004 and 1982/2004. Although these largely are designed just to clarify and standardise existing legislation, there are some pragmatic changes that increase asymmetries and especially some forms of transit trade. Regulation (EC) 1949/2005 brings Extrastat treatment of these areas in line with Intrastat treatment from Ships & Aircraft Stores Previously, national goods delivered (for either use on the journey by the crew or passengers, or for the operation of the craft) to foreign ships and aircraft were recorded as dispatches and similarly, foreign goods delivered to national craft were recorded as arrivals, as appropriate. However, now only the dispatch side need be recorded, leading to an asymmetry between Member States. This was a pragmatic decision due to the difficulty of correctly accounting for the import/arrivals side. Ships & Aircraft Previously, these were recorded as an import when there was a change in ownership from foreign ownership to a person established in that. Similarly an export was recorded when there was a change in ownership from an established person to foreign ownership. The new regulations now consider that an arrival or dispatch has occurred when a ship is transferred between the registers of two different countries. Gas and Electricity The imports and exports of energy goods like electricity and gas have caused specific problems due to their fungible nature. Thus a sale between entities in two different countries may not be directly traceable to a specific movement of physical goods. This is due to the pipeline or grid operator netting off the fiscal sale of goods across the border, and only physically pumping the net sale across the border. Worse, for sales that involve transit through an intermediate country, there is no guarantee of the true origin of the goods the buyer actually receives, since the entire product is mixed within the pipeline. Since Eurostat is primarily concerned with the physical movement of goods across borders, Regulation (EC) 1982/2004 allows the physical net flow across the border to be recorded as the arrival and dispatch between the two Member States, disregarding the fiscal sales behind this movement (including when such fiscal sales imply the transit of goods to another Member State). In effect, much of the trade seen in earlier years appears to be recorded as trade between the dispatch country and the country where these goods first cross a border, so there should not be a large change in the recorded trade figures. 24

25 9. Recommendations The current regulations regarding export declarations do satisfy the requirements of IMTS for exports via another country, whether that is another MS or not, as long as they are properly implemented. Some trader education however may be necessary to insure the proper completion of SADs and that no dispatch Supplementary Declaration (SD) is completed for the movement. The will issue guidance to such effect. directly from another country outside the EU causes no particular concern, there is however no easy solution for imports via another MS. The current freedom enjoyed by importers about how to declare their imports cannot realistically be changed purely for statistical purposes. It can be seen that the current system is pragmatic in that it avoids systematic asymmetries with in the EU, and each countries trade balance remains intact. Since no customs declaration (SAD) accompanies the arrival in the, all information has to be via the Intrastat SD. Some countries collect the optional country of origin field on SDs, which does allow them to reallocate transit trade from outside the EU for their national figures if they wish. However, as this field is not compulsory, it cannot be used for the trade statistics sent to Eurostat. There are several possible solutions to this problem: Include the Country of Origin as a compulsory field on SDs. However, current moves are to simplify Intrastat, not add more fields. There would also have to be an accompanying field on import SADs to identify the EU country of destination, so that Member States can uniformly identify imports cleared there but are destined for elsewhere in the EU. If this is not done, then when EU data are aggregated there would be double counting of EU imports. However, the SAD is also currently undergoing simplification. Rather than collect the country of origin data from all Intrastat traders, only collect this information from some of them. This could be achieved either by having the fields only being compulsory for those traders above a higher threshold (with minimum coverage defined by Eurostat), or by an additional survey of only the key traders involved in transit trade. While this would increase the burden on the provider, it would be substantially less then option one. Transit trade can be estimated by grouped analysis. However, this relies on trade data from other EU countries, so cannot be performed until several months after the period in question, and relies on knowledge of transit trade patterns so that the correct country groupings are used. Greater organisation in the sharing of such information, facilitated by Eurostat, would be useful. 25

26 10. Conclusion The correct attribution of partner country to transit trade is a longstanding problem. However, the problem has become worse over recent years, due in part to the separation of the Intrastat and Extrastat collection systems. There is the potential to improve the situation, but to eliminate the problems inherent in the collection systems (as opposed to those from trader error and ignorance) would unfortunately increase the burden on at least some businesses. 26

27 Annex A Procedure to be followed when a company in one European Union (EU) Member State exports goods to a non-eu country, via a second Member State for transport reasons only (Draft). The Rotterdam Effect 1. Introduction: a. The Rotterdam Effect is the term used to describe the potential statistical inaccuracy resulting from the misreporting of exports from the EU as Intrastat movements. This can happen where the goods pass through a port or airport in another EU Member State for transport reasons. Such goods are intended for export out of the EU and, typically, are merely transhipped in the second Member State from one vessel to another, prior to exiting the EU. For example: A company in Member State 1(MS1) loads goods onto a feeder vessel in the local port, which takes them to a port in Member State 2 (MS2). The goods are transhipped in the port of MS2 to a deep-sea vessel that transports the goods to a non-eu country, e.g. India. MS 1 company Feeder vessel MS 2 port MS1 export Deep-sea vessel India The goods are exported from MS1 to India, via a port in MS2. The MS1 company must therefore declare an export from MS1 to India, the country of final destination outside the EU. 27

28 2. Responsibility for export formalities: a. the export declaration should normally be lodged with customs in the actual Member State of export (MS1) who supervise the place where the exporter is established (see below), or where the goods are loaded for export shipment. Transhipment from a feeder vessel to a deep-sea vessel in an intermediate EU port does not constitute loading for export shipment. (See example 7a) b. an export declaration may only be lodged in another Member State (MS2) in exceptional circumstances, e.g. a change of contract or diversion from the original destination. (See example 7 c.) c. the Intrastat system must not be used to declare the removal of goods from one Member State (MS1) to an intermediate port or airport in another Member State (MS2) where the sale contract has been concluded with a company established in a non-eu country. 3. Export declaration: a. where goods are declared for export in MS1, but exit the EU via MS2, an endorsed copy of the Single Administrative Document (SAD) must normally accompany the goods to the office of exit (the port in MS2) to indicate that export formalities have been carried out in the country of export. b. where the copy SAD is not required to travel with the goods, the standard shipping note will normally be endorsed EXPORT by Customs. 4. Evidence of final destination: Evidence of final destination can normally be found in most of the following commercial documents: a. sales contract b. sales invoice c. the load list (where required) d. ship s manifest (these may only feature the identity of the deep-sea vessel with the goods destination). e. Bill of Lading 5. Exceptions: a. exceptions to this rule may include occasions where the transaction occurs between a company in MS1 and a middleman in MS2, even though the goods may simply be transhipped and exported from the EU. b. in this case the first trader may not know the final destination of the goods and will normally declare the transaction as a dispatch to the EU partner country. (See example 7 b.) 28

29 6. What is an established company? A company is regarded as established in a Member State if it has premises from which it engages in and intends to continue engaging in significant economic activities indefinitely or for finite, but long, period of time. A detailed definition can be found in paragraphs of the IMF Balance of Payments Manual. A company is not regarded as established solely because it is registered for VAT in a Member State. 7. Examples: a. export: Export Goods leave on a feeder vessel destined for India bill of lading covers carriage of goods from to India b. Intrastat/NL export: NL (Transit) Goods are transhipped in Rotterdam to a deep-sea vessel bound for India IN Import Goods arrive in India to fulfil an order placed with a supplier Dispatch NL Arrival IN Import Goods leave destined for a company company in the. declares arrival, partner company has no knowledge of country, subsequently final destination therefore arranges export to India dispatch, partner country NL c. export, destination changed in port of exit (NL) Goods arrive in India to fulfil an order placed with company Export NL (Transit) In Goods leave on a feeder vessel destined for India bill of lading covers carriage of goods from to India NL Export Goods originally destined as an export to India. Change of contract involves delivery of goods to Sri Lanka LK Import Goods arrive in Sri Lanka as an import from the (via NL) as a result of a change of contract 29

30 Annex B Import Asymmetries with and the, Millions Jan 1,159 1, ,771 1, Feb 1,309 1, ,891 2, Mar 1,553 1, ,282 2, Apr 1,392 1, ,912 2, May 1,507 1, ,046 2, Jun 1,381 1, ,080 2, Jul 1,377 1, ,013 2, Aug 1,268 1, ,062 2, Sep 1,353 1, ,246 2, Oct 1,437 1, ,240 2, Nov 1,476 1, ,281 2, Dec 1,233 1, ,002 2, Jan 1,583 1, ,889 2, Feb 1,514 1, ,807 2, Mar 1,612 1, ,360 2, Apr 1,407 1, ,785 2, May 1,552 1, ,966 2, Jun 1,543 1, ,261 2, Jul 1,568 1, ,903 2, Aug 1,373 1, ,802 2, Sep 1,526 1, ,034 2, Oct 1,531 1, ,092 2, Nov 1,476 1, ,076 2, Dec 1,209 1, ,731 2,

31 2002 Millions Jan 1,506 1, ,956 2, Feb 1,589 1, ,920 2, Mar 1,753 2, ,272 2, Apr 1,574 1, ,014 2, May 1,558 1, ,024 2, Jun 1,379 1, ,042 2, Jul 1,614 1, ,941 2, Aug 1,437 1, ,881 2, Sep 1,715 2, ,105 2, Oct 1,715 2, ,171 2, Nov 1,517 1, ,018 2, Dec 1,231 1, ,780 2, Jan 1,476 1, ,781 2, Feb 1,504 1, ,764 2, Mar 1,554 1, ,010 2, Apr 1,499 1, ,784 2, May 1,376 1, ,629 2, Jun 1,499 1, ,896 2, Jul 1,464 1, ,861 2, Aug 1,227 1, ,675 1, Sep 1,510 1, ,077 2, Oct 1,582 1, ,128 2, Nov 1,497 1, ,013 2, Dec 1,404 1, ,955 2, Source: Comext Database 31

32 Export Asymmetries with and the, Jan 1,239 1, ,751 1, Feb 1,347 1, ,819 1, Mar 1,510 1, ,256 1, Apr 1,115 1, ,753 1, May 1,465 1, ,977 1, Jun 1,470 1, ,996 1, Jul 1,170 1, ,890 1, Aug 1,316 1, ,036 1, Sep 1,341 1, ,130 1, Oct 1,302 1, ,899 1, Nov 1,289 1, ,415 2, Dec 1,149 1, ,088 1, Jan 1,430 1, ,976 1, Feb 1,237 1, ,825 1, Mar 1,328 1, ,037 1, Apr 1,084 1, ,701 1, May 1,309 1, ,936 1, Jun 1,440 1, ,077 1, Jul 1,216 1, ,902 1, Aug 1,120 1, ,898 1, Sep 1,213 1, ,938 1, Oct 1,219 1, ,931 1, Nov 1,222 1, ,922 1, Dec 1,070 1, ,686 1,

33 2002 Jan 1,217 1, ,743 1, Feb 1,323 1, ,863 1, Mar 1,451 1, ,083 1, Apr 1,325 1, ,909 1, May 1,429 1, ,799 1, Jun 1,337 1, ,840 1, Jul 1,441 1, ,746 1, Aug 1,107 1, ,561 1, Sep 1,343 1, ,969 1, Oct 1,358 1, ,785 1, Nov 1,326 1, ,843 1, Dec 1,176 1, ,780 1, Jan 1,282 1, ,923 1, Feb 1,386 1, ,872 1, Mar 1,404 1, ,165 1, Apr 1,388 1, ,635 1, May 1,287 1, ,470 1, Jun 1,389 1, ,562 1, Jul 1,317 1, ,542 1, Aug 1,140 1, ,210 1, Sep 1,285 1, ,433 1, Oct 1,277 1, ,458 1, Nov 1,229 1, ,446 1, Dec 1,203 1, ,499 1, Source: Comext Database 33

34 Annex C Import Asymmetries with and the by HS Chapter, Chapter Live animals 02 Meat and edible meat offal 03 Fish and crustaceans, molluscs and other aquatic invertebrates 04 Dairy produce; birds' eggs; natural honey; edible products of animal origin, not elsewhere specified or included 05 Animal originated products; not elsewhere specified or included 06 Trees and other plants, live; bulbs, roots and the like; cut flowers and ornamental foliage 07 Vegetables and certain roots and tubers; edible 08 Fruit and nuts, edible; peel of citrus fruit or melons 09 Coffee, tea, mate and spices Thousands 2,279 3, ,834 12,527 2,693 99, ,083 22, ,549 1,054,404 90,855 13,878 20,811 6,932 39,669 59,244 19, , ,700 19, , ,520 24,424 14,484 14, ,000 11,401 4,400 43,104 59,929 16,825 1,007,457 1,062,003 54, , ,487 4, , , ,603 76, ,898 27, , ,007 34,211 20,292 21,937 1,645 13,188 16,333 3, Cereals 11,445 34,251 22,806 44,809 41,902-2, Products of the milling industry; malt, starches, inulin, wheat gluten 11,279 9,502-1,777 30,987 25,953-5,033 34

35 Thousands Chapter 12 Oil seeds and oleaginous fruits; miscellaneous grains, seeds and fruit, industrial or medicinal plants; straw and fodder 13 Lac; gums, resins and other vegetable saps and extracts 14 Vegetable plaiting materials; vegetable products not elsewhere specified or included 15 Animal or vegetable fats and oils and their cleavage products; prepared animal fats; animal or vegetable waxes 16 Meat, fish or crustaceans, molluscs or other aquatic invertebrates; preparations 17 Sugars and sugar confectionery 18 Cocoa and cocoa preparations 19 Preparations of cereals, flour, starch or milk; pastrycooks' products 20 Preparations of vegetables, fruit, nuts or other parts of plants 21 Miscellaneous edible preparations 22 Beverages, spirits and vinegar 23 Food industries, residues and wastes ; prepared animal fodder 13,849 11,936-1,913 70,497 69, ,448 2, ,619 2, , , ,947 76,117 15, , ,456-89,918 71,997 81,112 9, , ,457 57,874 53,853 72,379 18,526 76,638 65,801-10,837 91, ,148 23, , ,967 4, , ,011 9, , ,708-2, , , , , ,966 90,066 75,065 88,483 13, , ,396-8,124 72, ,505 49, , ,858 96,234 68,744 48,575-20, , ,546 35,965 35

36 Chapter 24 Tobacco and manufactured tobacco substitutes 25 Salt; sulphur; earths, stone; plastering materials, lime and cement Thousands 4,310 12,599 8,289 66,259 64,418-1,841 10,994 18,258 7,264 14,946 25,141 10, Ores, slag and ash 18,508 5,750-12,757 28,030 37,339 9, Mineral fuels, mineral oils and products of their distillation; bituminous substances; mineral waxes 28 Inorganic chemicals; organic and inorganic compounds of precious metals; of rare earth metals, of radio-active elements and of isotopes 314, , , ,581 1,225, ,815 84, ,935 27, , ,553-9, Organic chemicals 533, ,913 64,892 1,305,642 1,198, , Pharmaceutical products 673,378 1,086, , ,538 1,232, , Fertilizers 22,083 31,931 9,848 67,804 64,039-3, Tanning or dyeing extracts; tannins and their derivatives; dyes, pigments and other colouring matter; paints, varnishes; putty, other mastics; inks 33 Essential oils and resinoids; perfumery, cosmetic or toilet preparations 123, ,871 21, , ,150 10,212 96,112 85,078-11, , ,630 33,663 36