52001DC0093

Report from the Commission to the Council and to the European Parliament on experience acquired in the work carried out pursuant to Council Directive 95/64/EC on statistical returns in respect of carriage of goods and passengers by sea /* COM/2001/0093 final */


REPORT FROM THE COMMISSION TO THE COUNCIL AND TO THE EUROPEAN PARLIAMENT on experience acquired in the work carried out pursuant to Council Directive 95/64/EC on statistical returns in respect of carriage of goods and passengers by sea

(Text with EEA relevance)

TABLE OF CONTENTS

1. Introduction

1.1. Background

1.2. Geographical coverage

2. Transitional period (Article 10)

2.1. Derogations

2.2. Pilot Studies

(a) feasibility and cost of collection of specific information

(b) technical aspects of data collection

3. Ports (Article 4)

4. Accuracy of statistics (Article 5)

5. Processing and Transmission of the results of the data collection (Articles 6 and 7)

5.1. Raw data collection and processing by CNA

5.2. Transmission of the results to the Commission

6. Dissemination of statistical data (Article 9)

7. Financial contribution (Article 11)

8. Detailed rules of implementation and procedure (Articles 12 and 13)

9. Implementation (Article 14)

9.1. Belgium

9.2. Denmark

9.3. Germany

9.4. Greece

9.5. Spain

9.6. France

9.7. Ireland

9.8. Italy

9.9. Netherlands

9.10. Portugal

9.11. Finland

9.12. Sweden

9.13. United Kingdom

9.14. Norway

9.15. Iceland

10. Conclusion

1. Introduction

1.1. Background

On 8 December 1995, the Council adopted a Directive on statistical returns in respect of carriage of goods and passengers by sea. The Directive 95/64/EC establishes a harmonised framework for the collection of maritime statistics across the European Economic Area. It defines the information to be periodically transmitted from each Member State to the Statistical Office of the European Communities, for the carriage of goods and passengers by sea.

Article 8 of the Council Directive stipulates that the Commission shall submit a report to the Council on experience acquired in the work carried out pursuant to this Directive after data have been collected over a period of three years. Data collection started with reference year 1997.

Where no other legal act is cited in this report, all articles refer to Council Directive 95/64/EC [1], hereafter referred to as "the Directive" or "the maritime Directive". The Commission adopted two subsequent legal acts laying down implementing rules for the Directive: Commission Decision 98/385/EC of 13 May 1998 and Commission Decision 2000/363/EC of 28 April 2000 [2]. The main forum for all discussions and decisions concerning the implementation and application of the Directive is the EEA Working Group on Maritime transport statistics, hereafter referred to as "the Working Group".

[1] OJ L 320, 30.12.1995, p. 25.

[2] OJ L 174, 18.6.1998, p. 1, and OJ L 132, 5.6.2000, p. 1.

The maritime Directive aims at completing the information available on transport by other modes at European level - road, rail, inland waterways, and aviation - for enabling the European Commission to carry out the related policies.

1.2. Geographical coverage

The Directive is applied by all EU Member States except Luxembourg and Austria, which do not have maritime ports on their territory. Besides these thirteen countries, the two EFTA countries that are member of the European Economic Area (EEA) and possess maritime ports, Iceland and Norway, are associated to the maritime information system of the Directive.

The procedure for integrating the maritime Directive in the Agreement on the European Economic Area was accomplished on 6 March 1998 with the adoption of Decision 17/98 by the EEA Joint Committee. [3]

[3] Decision 17/98 entered into force on 7.3.1998 and was published in OJ L 272, p. 24 on 8.10.1998.

The Decision amended Annex XXI on statistics of the EEA Agreement so as to include Council Directive 95/64/EC in a newly introduced point 7B. Two adaptations to the Directive are made for the purpose of the EEA Agreement, concerning small Norwegian ports and codes for Norwegian vessels (Annex V of the Directive). Commission Decision 98/385/EC was also integrated in the EEA Agreement.

Due to a procedural modification, the participating ports of the EEA/EFTA countries are no longer laid down by Commission Decisions, but by modification of the EEA agreement. This reflects the wish of these countries of being fully involved in the decision making process. Thus Commission Decision 2000/363/EC revised the list of European ports without including the Icelandic and Norwegian ports.

2. Transitional period (Article 10)

2.1. Derogations

The partners in the European Statistical System (ESS) agreed on starting the regular data collection one year after the entry into force of the Directive, on 1 January 1997, and to qualify the first three years as a transitional period. This period should enable Member States to adopt all necessary administrative and legal measures for implementing the Directive at national level.

During this period, each Member State was granted derogations that reflected its specific needs for adaptation. Derogations consisted either in excluding individual ports from the collection duty, allowing the use of national codes, or providing exceptions for specific variables. The number of derogations was reduced after the first year of implementation, assuming that Member States would already have made progress in adapting their collection systems.

On the one hand side, these derogations were necessary, as setting up the data collection system required efforts and resources. On the other hand side, as a consequence, data collected during the first three reference years are rather incomplete.

Since 1 January 2000, all Member States have to provide complete data sets, classified and coded in accordance with the requirements of the Directive.

2.2. Pilot Studies

In line with Article 10(2) of the Directive, the European Commission adopted a programme of pilot studies during the transitional phase, covering two domains:

(a) the feasibility and cost of collection of specific information, and

(b) technical aspects of data collection.

Funding for these pilot studies mainly stemmed from the Fourth Framework Programme for Research and Development, in particular from the SUPCOM and IDA/DSIS (Interchange of data between administrations/Distributed statistical information service) frameworks.

(a) feasibility and cost of collection of specific information

When the Directive was drafted, the collection of certain additional items of information within the regular collection system were considered. As no agreement on this could be reached, it was decided to study the feasibility and cost to the Member States of collecting such information on a regular basis through a series of pilot studies. The information concerned was:

i description of the goods transported

ii passengers transported over short distances

iii information concerning feeder services and transport chains

iv data relating to the nationality of the maritime transport operator

For items i, iii and iv, the studies' purpose was to determine whether the current system could be extended by these additional variables. For item ii, the scope of the study was rather a reduction than an extension of the current system.

The pilot studies were carried out by the Commission and funded under the fourth framework programme. Their results were discussed by the Working Group on maritime transport statistics in June 1999. Given the difficulties of implementation of such new items, the Group decided not to launch any procedure for extension of the Directive immediately, but to wait until the regular collection system is well established. The Group created a Task Force that should further explore the issue. In parallel, the Commission continued to develop a detailed methodological framework, among other subjects on passengers transported over short distances.

The main results of the studies are:

i description of the goods as defined in Annexes III and VIII, data set B1

As the current setting of the Directive foresees only type of cargo information, detailed comparison with statistics on inland transport mode are not possible, as these use commodity information according to the NST/R 24-group classification.

The pilot study therefore explored the cost and feasibility of collecting commodity information in ports according to this classification.

Such detailed information appears to be feasible and at reasonable cost to be collected for bulk and semi-bulk cargo.

The main difficulty and cost however occur in compiling such analyses for container and ro-ro traffic, in most Member States. Collecting information on container content would require high resources, as substantial coding of text description would be necessary. Also the high number of different consignments in one container was found to increase cost.

For ro-ro traffic, the transport documents in general do not contain usable commodity information available to the operators.

Member States collect commodity information at various degrees of detail. Any classification chosen should be common to other transport modes, and potential revisions of the NST/R classification should be taken into account.

ii passengers transported over short distances

The Directive foresees collection of data on passengers transported by sea without providing any minimum distance of the journey. Theoretically, also short trips should therefore be recorded. As some countries do not collect data below a certain distance, there may be distortions in the common database. The scope of the pilot study was to consider the possibility for Member States to exclude from the survey part or all of the minor routes that are undertaken in their country.

Nine Member States already cover all passenger movements and two will do so starting in 2000. The outcome of the study suggests that Member States do not have substantial problems in obtaining all passenger information. On the contrary, it might increase the burden on their data providers if new thresholds for excluding minor passenger movements were created, and thus be counterproductive. Collection of passenger data might benefit from the information provided in connection with Council Directive 98/41/EC of 18 June 1998 on the registration of persons sailing on board passenger ships operating to or from ports of the Member States of the Community (OJ L 188, 2.7.1998, p. 35).

It was found that the length of a journey, i.e. the number of miles over which passengers are transported, is not necessarily a meaningful indicator for the importance of a certain ferry service. The size of the flow may be more relevant: Through the threshold of 200 000 passengers per year (Article 4 of the Directive), traffic flows of minor importance are already excluded. The study also pointed at the problem of some countries applying the criterion "on seagoing voyages" (Article 2): where this is the case, very short routes are excluded from passenger statistics, even where such services carry a large number of passengers.

iii information concerning feeder services and transport chains

In response to the increasing demand for information on intermodal chain data from the user side, the feasibility and cost of collection of such data was explored by a pilot study. The findings were that this information is not available from the sources used for the Directive. The only possibility would be to collect data from ports, for which it would be a relatively small burden if they seize the data already for pricing purposes. Several delegations agreed that some information is available in ports as they charge different prices. In some countries however, this may be very costly and/or could only be done through separate sample surveys or from other transport modes.

The Working Group gave a mandate to the Task Force on extension to discuss methodological aspects of feeder services and to co-ordinate with the Intermodal Expert Group.

iv data relating to the nationality of the maritime transport operator

The pilot study on obtaining information on the nationality of the maritime transport operator reflects a stated need of users for information on the main economic beneficiary of maritime operations and on the European fleet in general. The study's outcome was that collecting the information required by the Directive - i.e. the real centre of commercial control - is not feasible. The main difficulty was seen in defining the country of economic benefit, as ownership and operation are extremely complex. The cost for any such data collecting would be high and was not seen as justified. The Working Group considered exploring alternative methods, such as consulting national associations or consulting commercial databases that could use a complex algorithm for defining the country of economic benefit.

(b) technical aspects of data collection

A set of pilot studies was established for studying the possibility of collecting data under the arrangements concluded in the context of the simplification of trade procedures, the International Standards Organisation (ISO), the European Committee for Standardisation (CEN) and international customs rules (Article 10(2)(b)).

The studies mainly contributed to the EDIMARS (electronic data interchange in maritime statistics) project launched in 1995. The first trials were carried out in Spain, Netherlands and France, followed by Ireland, Germany and the UK and later all other Member States. The project was open to all actors of the maritime transport sector within the European Union: national and regional administrations, port authorities, transport operators, ship owners, ship agents, carriers, port community systems providers, etc.

In the first three years, the project focused on the collection of raw data from the providers i.e. ports, carriers or ships agents. In the fourth year, it mainly aimed at increasing the use of electronic means by the competent national authorities for sending data to Eurostat.

The EDIMARS project facilitated the exchange of experience in the organisation and implementation of maritime statistics data collection. It helped reduce the cost and improve the efficiency of collecting and disseminating maritime transport statistics, through the use Electronic Data Interchange (EDI). EDI is commonly defined as the application-to-application transfer of business documents between computers. Many businesses choose EDI as a fast, inexpensive, and safe method of sending purchase orders, invoices, shipping notices, and other frequently used business documents. As traditional means of collecting information (queries, interviews, polls etc.) have proven expensive, not exhaustive, and not so reliable, Eurostat and the European Board for EDI Standards (EBES) have designed a common language for exchanging statistics. GESMES, the Generic Statistical Message, uses the EDIFACT standard (Electronic Data Interchange for Administration, Commerce and Transport).

The status quo at the beginning of the project was a large number of different formats (proprietary formats, Gesmes) and supports (paper, diskettes, magnetic tapes, e-mail) being used. By March 2000, no Member State used manual methods like paper anymore, some still used a mixture of diskettes and EDI transmissions, and some countries like Spain, Portugal and Finland systematically use EDI transmissions.

The project demonstrated through real trials that the use of EDI for collecting maritime statistics brings two major advantages:

-a homogeneous language for transmitting the data sets required by the Directive using GESMES for maritime transport statistics;

-there are ways of automatically extracting the statistical data from operational EDIFACT messages such as IFCSUM (International Forwarding and Consolidation Summary Message) or CUSCAR (Customs Cargo Report), as well as from Port Community Systems or Information Systems.

Documentation produced within the project provides a complete description of how to extract the statistical elements required in the Maritime Directive from an IFCSUM message.

In parallel to these pilot studies, international and European groups developed standard messages, such as EDIMAN, Maritime Cargo Manifest Message, which is based on the UN/EDIFACT syntax and was developed by the Manifest Group of the Maritime Commission. The Commission followed the work of these Groups, including the International Transport Implementation Guidelines Group (ITIGG) and the UN/CEFACT Codes Working Group, which is responsible for maintaining the set of UN/ECE Recommendations covering nomenclatures to benefit global trade facilitation. This Group deals with revisions of modes of transport codes, package type codes, commodity coding for transport purposes, types of means of transport and types of cargo and UN/LOCODES, which are particularly important for maritime statistics.

The EDIMARS project also contributed to the work of this group by adopting the ITIGG CUSCAR rules for the third year pilot trials in the UK and in Germany. ITIGG is currently extending its work to cover the other modes of transport, which may offer great benefits not only to the other modes for harmonised raw data collection but also for the collection of intermodal statistics.

3. Ports (Article 4)

As reporting units, ports play the key role for data collection within the information system of the Directive. According to Article 1, Member States shall collect Community statistics on the carriage of goods and passengers by seagoing vessels calling at ports in their territories.

In the 1993 draft of the Directive, the criteria for establishing a list of ports was that the list should at least cover, for each Member State, 90% of all sea transport of goods and 90% of all movements of passengers by sea. The final act however stipulates merely that a list of ports should be drawn up, without providing objective criteria.

In line with the procedure foreseen in Articles 4(1) and 13, a first list of European ports was applied since 1997 and formally adopted in 1998 with Commission Decision 98/385/EC. It comprised a total of 1575 ports in all 15 EEA countries, out of which 1302 were statistical ports. Starting with the first year of full application, 2000, a new list of ports has been applied. It was formally adopted by Commission Decision 2000/363/EC and comprises a total of 1357 ports in 13 EU countries, out of which 1089 are statistical ports (see table 1). Around 160 EEA/EFTA ports will be added to the list through a separate procedure.

The selection was carried out by each Member State for the ports on its territory. The number of ports, the relative number of ports per kilometre of coastline and also the grouping of several sub-ports to a national statistical group differ considerably from country to country. This is due to local port administration structures and national traditions in sea transport and also maritime statistics. There is no common minimum number of passengers, ship movements or cargo handling per year, for a port to be included in the list.

In the United Kingdom, for instance, the total number of ports in the list is double the number of ports that actually provide statistics as a national statistical port. In Ireland and Portugal, on the contrary, each port listed is also a statistical port. In Italy, where all ports are systematically included, each of them reports as a separate statistical port, without aggregation.

According to the Directive, a port is defined as a "place having facilities for merchant ships to moor and to load/unload cargo or to embark/disembark passengers to or from vessels". In exceptional cases, Member States have chosen a whole navigable way as a statistical port, as is the case for the Belgium Albert channel.

Only the analysis of collected data can show whether the heterogeneity of the lists is an obstacle to the production of comparable and accurate information on EU maritime transport.

Among the ports included in the list, only those exceeding a certain annual threshold in cargo or passenger transport need to report in detail and on a quarterly basis. The ports below this threshold only need to provide summary data on an annual basis. The threshold is fixed at one million tonnes of goods or 200 000 passengers per year. During the transition period from 1997 to 1999, Member State may make use of a derogation foreseen by the Directive, and apply the provisional threshold of two million tonnes or 400 000 passengers per year. The number of ports on which data are contained in the common database vary according to the data set.

Due to this threshold restriction, the derogations granted and the revision of the port list, meaningful time series for the majority of ports can only be started from reference year 2000 onwards.

Table 1 Number of ports in which maritime statistics should be collected

>TABLE POSITION>

From experience with data transmitted by the Member States, it appeared that only a part of the ports that have a reporting duty actually declare traffic. A substantial revision of the port list may become necessary in the future, when the information system is well established.

4. Accuracy of statistics (Article 5)

Intra-Community transport flows have the particularity of being recorded twice: at the origin and the destination of the flow. The data collection characteristics of the maritime Directive include information on relation (i.e. the port of loading/unloading for the EEA, maritime coastal area for non-EEA) both for passengers and goods consignments. It thus enables the building of mirror tables, which can be used for quality control. Differences in declarations between two reporting units clearly indicate methodological problems.

At country-to-country level, differences in total declarations may stem from the fact that only one of the two concerned ports of a traffic flow is included in the port list, while the port of (un-)loading does not have a reporting duty. At port-to-port level, in turn, discrepancies are easy to identify.

The issue is being discussed in the Working Group since June 1999; the Commission proposes a bilateral approach: two Member States with high differences in reporting on an identical flow should be notified by the Commission and try to compare their methodologies in order to identify and remove or reduce the source of the discrepancies. Such bilateral practices have proven successful in other domains such as migration. Even if the deviation cannot be removed completely, the research may help building factors for estimations.

Portugal and several scandinavian countries have already started studying the coherence of data at national level, detecting a number of reasons for discrepancies:

-flows were declared in different months as they started at the end of one month and were finished at the beginning of the next month

-goods leaving one port destined for another; delays en route may introduce distortions vis-à-vis the reported data for both ports originally involved

-different statistical sources have been used: ship's manifest vs. knowledge of the vessel

-certain ports use specific unit equivalences for certain types of goods (for instance tonnes = cubic metres)

The Working Group has begun to undertake a series of continuous activities for improving the quality of the collected statistics. There is a number of methodological issues that require further research: such issues may concern national specificities or general problems such as determining the contents of containers, information on empty containers or micro-cabotage.

5. Processing and Transmission of the results of the data collection (Articles 6 and 7)

5.1. Raw data collection and processing by CNA

As far as raw data collection and processing of the collected data by the competent national authority (CNA) is concerned, the Directive only binds Member States as to the result, which is to produce comparable statistics with the agreed standard of accuracy. The choice of processing tools and work organisation is done at national level, according to the principle of subsidiarity. Some countries continued to apply their existing processing systems, adjusting them to the requirements of the Directive where necessary, while other countries introduced complete new systems. The financial contribution towards the cost of implementation was partly used to fund this set up of adequate processing environments.

The findings of the pilot studies concerning data collection are summarised in the chapter on Transitional period: pilot studies of this report. A project underway in the UK in 1999/2000 is piloting five different methods of raw data collection. The UK pilot projects comprise amongst others the SDES (Statistical Data Entry Software) and a set of data entry web forms for data providers.

5.2. Transmission of the results to the Commission

For the transmission of the collected data to the Commission, in turn, the Directive takes a more harmonised approach. It predefines the periodicity and structure of the data sets to be transmitted. The technical details for transmission shall be laid down by a Comitology procedure.

Given the diversity of tools used in the Member States, and the considerable adaptation effort necessary, the Commission did not propose common technical standards in the first phase of application of the Directive. Instead, the Commission supported Member States' efforts to install efficient transmission tools by funding pilot projects and developing a toolbox of software tools in generating electronic submissions to Eurostat. The aim of the Commission's EDI strategy is to allow some choice but at the same time to restrict the options to just two, both of which can be checked and processed within the Commission without having to be reformatted or manually edited.

The first out of the two proposed methods is transmission via a subset of the UN/EDIFACT GESMES message as defined by the "Maritime GESMES message implementation guidelines", which were developed during the EDIMARS project. This "Easy Maritime GESMES EDI solution" has been tested in practice and constantly improved.

The second method consists in sending a structured flat-file in comma separated format (".csv") following a structure, which was developed by the Commission for enabling automatic validation. Its concept is similar to the GESMES structure but it is a simple flat-file format. Both comma (,) and semicolon (;) can be used as separators and datasets can be accepted with or without empty fields. The structured flat-file solution is a no-cost solution, which can be directly derived from common data processing applications.

The flat-files or GESMES files can be sent either via STADIUM or by structured email attachments to a pre-defined functional E-mail address.

The overall design criteria of the "MAKEDISI EDI Toolbox" were maximum simplicity, portability across platforms and also 'full' EDI (Electronic data interchange) compatibility, i.e. not requiring any human intervention at all. In addition, the tools should be put at Member States' disposal without charge; they should be independent of software vendor or application, parameterised in order to easily allow amendments of code lists or GESMES; and should work in conjunction with standard communications software such as email packages. The PERL language was chosen, as it not only satisfied the design criteria but also because it offers additional functionality suitable for structured flat file processing.

The toolbox also provides the functions of format and code validations and of conversions between structured flat-files and GESMES, thus adding to the accuracy of data.

The "MAKEDISI EDI Toolbox" accepts both input formats: Maritime GESMES and structured flat-files. The only output format is GESMES, in order to move towards 100% GESMES transmissions, which corresponds to the general transmission policy within the European Statistical System.

A Commission managed newsroom on the CIRCA web-site provides support to the users of these tools and for sharing toolbox experiences, wish lists, frequently asked questions etc.

6. Dissemination of statistical data (Article 9)

The Commission has planned to include the collected data in the regular dissemination scheme of the European Statistical System, which consists in public dissemination on a number of supports to all public and private users, and privileged dissemination to the data providers and Commission Directorate General for Transport and Energy. The main technical environment for the dissemination of maritime data will be theme seven of the online database NEW CRONOS. CNAs are granted a free access, while the general public can purchase extractions from it.

In addition, dissemination of selected data on CD-Rom, or on Commission paper publication series such as Statistics in focus, yearbooks and Panorama on transport is foreseen.

This publication scheme responds to the expectations of the CNAs expressed in a questionnaire in 1998 and in subsequent Working Group meetings.

The main obstacles to dissemination during the transitional period were on the one hand that according to the Directive, the arrangements for dissemination need to be laid down by a Comitology procedure, and that there has not yet been a common agreement in the Working Group on the level of detail in which data should be disclosed. On the other hand, most Member States allow the dissemination of their data only under the condition of reciprocity, i.e. that the same data must be available for all countries at the same time. The situation as regards data reception in the Commission does not allow for such a complete publication, as several countries have not yet provided data. Even if all countries had transmitted the data according to their reporting duty, the derogations granted during the first three years of application of the Directive would entail considerable differences between countries when building tables.

In its 1999 meeting and by written follow up, the Working Group reached a provisional agreement on dissemination, which consists in publishing data between ports and maritime coastal areas (MCA). It was also agreed to review this agreement with a view to increase the level of detail in which data may be published after some experience is acquired.

These agreements do not give an indication on what should be published, but on what may be published without prejudice to the confidentiality requirements of the Directive and of related legal acts. The Member States are currently screening the concerns about commercial confidentiality among their respective data providers. First feed back from some countries indicated different views on the side of the providers, depending on the country of operation but also on the size of the reporting unit (small ports or operators tend to find information on their activities more sensitive than large ports or operators).

The formalisation of the Working Group's agreement in a form of a Commission Decision is under way.

7. Financial contribution (Article 11)

Article 11 of the Directive foresees that Member States shall receive a financial contribution towards the cost of work they carry out for implementing the Directive during the first three years.

The costs of implementation that occurred in the competent national authorities were substantial: total costs declared by all Member States in the two budgetary exercises 1998 and 1999 summed up to 3.61 million Euro.

Table 2 Overview on EU contributions

Member State // EU contribution (1998+1999)

BE // 66 000

DA // 128 000

DE // 150 000

EL // 82 000

ES // 50 491

FR // 128 000

IE // 38 000

IT // 167 735

NL // 58 000

PT // 20 000

FI // 16 000

SV // 37 000

UK // 194 224

Total // 1 135 450

In line with the provisions of Article 11(2) and (3), the European Community has contracted with Member States to contribute about one third of these costs, with 1.14 million Euro. The work that was taken into consideration concerned both adaptations within the competent national authorities and developments of software tools that would be distributed by the competent national authority to their ports or other data providers.

The costs occurring varied considerably between Member States, depending mainly on the previous state of data collection on maritime transport in each country. In some countries, the implementation of the Directive was undertaken in a joint effort to modernise the existing system of data collection.

In addition to these allocations towards Member States' costs of implementation, the European Commission has used financial resources for implementation within Eurostat and for the pilot projects mentioned in chapter Transitional period.

8. Detailed rules of implementation and procedure (Articles 12 and 13)

All detailed rules for implementing the Directive are to be laid down by a management Committee procedure specified in Article 13. So far, two Commission Decisions containing a set of detailed rules were adopted for implementing the Directive. They included a first list of ports and as well as a first update of this list. Furthermore, they completed and adapted the annexes of the Directive.

The port list and the other annexes with data collection characteristics will regularly be reviewed for reflecting the changes in the tonnage throughput and passenger transport of the ports, and for adapting them to economic and technical developments in general. (Ports whose tonnages fall below the threshold in some years may be maintained in the selected port list.)

The procedure foreseen by the Directive requires a certain amount of time for the final rule being formally adopted; in questions where all partners involved in the statistical process agree, it may therefore be conceivable to adopt a working solution rather than formalising each rule immediately. This has been the practice for detailed rules for transmission, where the Working Group agreed to apply the options proposed by the Commission on the basis of a gentlemen's agreement.

9. Implementation (Article 14)

Article 14 requires that Member States adopt the laws, regulations and administrative provisions necessary to comply with the Directive by 1 January 1997 and communicate them to the Commission.

By the end of the transitional phase, in December 1999, all Member States except Ireland have communicated the legal and administrative measures they took for implementing the Directive. The Commission considered the derogations granted during the transitional phase as a justification for delaying full implementation until that date.

The detailed situation concerning legal and administrative implementation in the Member States, derogations granted under Commission Decision 98/385/EC and data transmitted is set out below. It reflects the situation as at November 2000 and is based on experience with data reception in the Commission and on reports made by national delegations at the meetings of the Working Group and of the Co-ordinating Committee on Transport Statistics.

9.1. Belgium

Most of the information required by the maritime Directive was already collected by the Benelux declarations 20 and 21 on the basis of a Ministerial decree of 1975. [4] For collecting the remaining items of information, a Circular of the Customs and Excises Administration of 1997 [5] has introduced a supplementary declaration form for container transport and ro-ro-units that is annexed to declarations 20 and 21. The Circular contains an explicit reference to the Council Directive.

[4] "Arrêté ministériel du 13/11/75 relatif à la déclaration générale en matière de douane à l'entrée et à la sortie de navires." Published in Moniteur belge of 11.12.1975, p. 15811-15818. In force since 1.1.1976.

[5] "Circulaire du 1.2.1997 de l'Administration des douanes et des accises D.D. 92.791 (abroge la Circulaire du 1.1.1995 n° D.L. 1/7.887)". This declaration form replaces an earlier form introduced by Circular No. D.L. 1/7.887 of 1.1.1995 which entered into force on 1.1.1997.

Collection of all information required by the Directive should be carried out since 1 February 1997. Belgium was granted derogation for the port of Antwerp for relation and port of loading/unloading.

Belgium has transmitted all quarterly and annual data for reference years 1997 to 1999, including corrections. The delay of data transmission was continuously reduced over the years, and the last annual data sets were transmitted in accordance with the time frame set by the Directive.

9.2. Denmark

Denmark has already been collecting data on maritime transport before the adoption of the Council Directive. The collection was carried out in two surveys: one on ferry transport of passengers and goods and another one on throughputs in all ports. In order to comply with the Directive, a third survey was introduced, covering the transport of goods in major ports, starting in 1997.

Statistics Denmark is carrying out the data collection under the general authorization of Act no. 196 of 8 June 1966. [6] A specific legal act was not required for transposing the Directive.

[6] As subsequently amended (most recently by Act no. 1025 of 19.12.1992).

Denmark has transmitted all quarterly and annual data sets for reference years 1997 to 1999, and for first quarter 2000 to the Commission. Initial technical problems with codes in data of the first quarters were resolved. The delay of data transmission was reduced over the years, and the last annual data sets were transmitted in accordance with the time frame set by the Directive.

9.3. Germany

Germany has been collecting statistical information on maritime transport since 1957. In order to implement all variables and classifications foreseen by the Directive, the legislation in force [7] was amended on 17 December 1999 [8]. The amendment contains an explicit reference to Council Directive 95/64/EC and is in force since 1.1.2000.

[7] "Gesetz über die Statistik der Seeschiffahrt" of 26.7.1957, in force since 26.8.1957, published in a consolidated version in BGBl Part III, No. 9510-4, and "Verordnung über die Meldestellen für die Seeverkehrsstatistik" of 24.4.1958, in force since 1.5.1958, published in BAnz No. 80, last amended by decree of 5.11.1992 (BAnz p. 8761).

[8] "Gesetz zur Neuordnung der Statistiken der Schifffahrt und des Güterkraftverkehrs", published in BGBl of 22.12.1999, Part I, No. 56, p. 2452.

As Germany has been granted a large set of derogations [9] by Commission Decision 98/385/EC, it can comply with the Directive during the transition period on the basis of its 1957 law. The derogations concern passenger data and the use of the Directive's nomenclature for all variables except the reporting port and size of vessel.

[9] For all data sets partial derogation for port of loading, relation (MCA), type of cargo; total derogation for number of passengers in A3, D1 (nationality of vessel); partial derogation for nationality of vessel in E1 and type of vessel in F1 (all throughout the transition period).

Germany has transmitted quarterly and annual data sets for all reference years to the Commission, however the data are coded and classified according to the national system. As this system differs from the nomenclature defined by the Directive, the transmitted data sets, after applying a filter made by Eurostat, were integrated in the Eurostat production database in an aggregated format. Annual data sets and those of third and fourth quarters were transmitted in time.

9.4. Greece

For complying with the Directive, Greece has adopted two legal instruments: a Common Ministerial Decision on the Command and authorisation of the implementation of specific statistical works in the field of maritime transport of passengers, goods and vehicles" [10], applicable since 1 January 1997, and a Common Permanent Circular on the Implementation of specific statistical works and harmonisation of the statistical registration of maritime transport of passengers, goods and mobile units (vehicles), in accordance with Directive 95/64 of the European Union, by the Department of Merchant Marine Statistics [11], applicable since 1 January 1998.

[10] "ÁíÜèåóç êáé Ýãêñéóç äéåíÝñãåéáò åéäéêþí óôáôéóôéêþí åñãáóéþí óôïí ôïìÝá ôùí èáëÜóóéùí ìåôáöïñþí åðéâáôþí, åìðïñåõìÜôùí êáé ôñï÷ïöüñùí. "No. 6792/G-81 of 11.4.1997; OJ 333 of 23.4.1997.

[11] "ÄéåíÝñãåéá åéäéêþí óôáôéóôéêþí åñãáóéþí êáé åíáñìüíéóç ôùí óôáôéóôéêþí êáôáãñáöþí èáëáóóßùí ìåôáöïñþí åðéâáôþí, åìðïñåõìÜôùí êáé êéíçôþí ìïíÜäùí (ôñï÷ïöüñùí), óýìöùíá ìå ôçí Ïäçãßá 95/64 ôçò ÅõñùðáúêÞò ¸íùóçò, áðü ôï ÔìÞìá Óôáôéóôéêþí ÅìðïñéêÞò Íáõôéëßáò". Circular of 15.4.1998.

In parallel, a new computer system was set up, and a new methodology and a separate questionnaire for registration of passengers, mobile units and goods for domestic and international lines were developed.

Greece was granted extensive derogations, only summary information was required during the transitional phase (no information was required on relation, vessel details, and type of cargo (in some data sets)). Freight data collection started in 1998. In practice, Greece experienced problems with non-response of data providers, which led to delays in building the data sets.

Greece has only transmitted general cargo and passenger data for reference year 1997.

9.5. Spain

Spain has already been collecting maritime transport statistics before the adoption of the Council Directive and did therefore not ask for any derogation. The legislation in force enables the Spanish authorities to collect and transmit data in accordance with the Directive without any further modification. For implementing the Directive, a second version of the computer application SIGMA was developed, thus incorporating all variables, classifications and codes required by the Directive. In practice, a methodological problem occurred in collecting the port of loading/unloading, which seems to be often confused with the origin or final destination of cargo transported.

Spain was one of the first countries to transmit data to the Commission, and has provided all quarterly and annual data sets for reference years 1997 to 1999, with improving timeliness of transmission. However several major Spanish ports such as Bilbao and Valencia were not included in the provided data sets.

9.6. France

The legislation in force was sufficient for collecting data in the first two categories of ports in France: the seven largest ports, which are public enterprises, and the large ports of national interest. For prescribing data collection in accordance with the maritime Directive in the third category of decentralised ports, which have to provide statistics under a decree of 8 August 1986, an amending ministerial decree was adopted on 28 December 1999 [12].

[12] Published in the OJ of 31.12.1999 and in the Official Bulletin of Ministère de l'Equipement, des Transports et du Logement on 10.01.2000. The annexes were published in the Official Bulletin on 25.02.2000.

The Ministry of Transport has provided an information system to several ports allowing them to monitor transport movements in the ports. This software (TRITON) is adapted both to the requirements of the Directive and to the internal modernisation programme. For those ports equipped with other software, the Ministry has reminded them of the obligation to adapt their systems to the output file structures of the TRITON application. At the same time, the Ministry has installed an Intranet site that will be accessible to all French maritime services by the end of 2000.

In practice, France experienced some problems with the mobilisation of respondents and the data flows from ports to the national administration.

France was granted partial derogation for information on type of cargo for all data sets and total derogation for nationality of registration.

France has only transmitted the annual data set A3 for 1997 and for 1998, and all data sets for the first quarter 2000.

9.7. Ireland

Ireland has so far been collecting maritime transport statistics under the Statistics Act of 1993. A supplementary Statutory Instrument is currently being drafted by the Parliamentary Draftsman's Office. This instrument will make the provision of maritime data from ports (per Maritime Directive) a statutory requirement.

Until its entry into force, data are collected on a voluntary basis according to the Agreement with the Ports Working Group (established in 1996) to provide the required statistical data. In co-operation with the ports that are beyond the threshold of the Directive (and therefore have to report detailed data), CSO developed a central manifest to be completed by the shipping operators. For small ports, which have to report less detailed data, a second survey was established. The new survey "National Survey of Ports in Ireland" is ongoing since the first quarter of 1997.

Ireland had asked for a limited number of derogations only, in particular for type of cargo information. The Commission received all quarterly and annual data sets for reference years 1997 to 1999, except for the port of Rosslaire. Due to some code problems occurring during the integration into the common database, all data sets were transmitted again.

9.8. Italy

The survey on maritime transport statistics carried out in Italy since 1995 covered some of the requirements of the Directive. According to the Italian legislation, the survey is mandatory for respondents.

For meeting the remaining requirements, the data collection model needed to be adapted. For this purpose, Italy adopted an administrative measure introducing a new, fully revised questionnaire on statistical returns on maritime transport. In practice, Italy experienced technical and administrative difficulties with this new model which was applied from 1 January 2000 onwards.

Due to derogations granted by Commission Decision 98/385/EC, the Directive has to be fully applied only from January 2000 onwards. Only direction, number of vessels, tonnage and unit load needed to be provided according to the nomenclature defined by the Directive during the transition period.

Italy has transmitted quarterly and annual data sets (except A2 and C1 for which there was a derogation) for reference years 1997 and 1998.

9.9. Netherlands

The Netherlands have already been collecting maritime transport statistics before the adoption of the EC Directive. For implementing the Directive, some technical adaptations and an amendment of the Dutch national law were necessary. The amendment was made through Decision of 5 July 1997 [13], which contains an explicit reference to the EC Directive.

[13] "Besluit van 5 juli, houdende vaststelling van bepalingen met betrekking tot de verstrekking van scheepvaartgegevens voor statistische doeleinden (Besluit statistische gegevens scheepvaartverkeer)", published in "Staatsblad van het Koninkrijk der Nederlanden, No. 341 of 5.7.1997.

During the three-year transitional period, the Netherlands were granted partial derogation for information on "port of loading/unloading", "relation" and "type of cargo".

The Netherlands provided all quarterly and annual data sets for the reference years 1997 to 1999 to the Commission. The time frame of the Directive was usually observed for the fourth quarters and annual data. In line with the derogation, these data are broken down by country of loading/unloading, not by port or MCA. The data were integrated in the common database after translation of national codes. All data sets for the first quarter 2000 were transmitted using the Directive's codes.

9.10. Portugal

Thanks to its tradition in collecting maritime statistics, Portugal did not need to ask for any derogation from its reporting duties under the maritime Directive. The Statistical law in force [14] enables the Portuguese National Statistical Institute to collect and transmit data in accordance with the Directive without any further amendment. Data are collected in a new survey since 1997 on a monthly and annual basis.

[14] "Lei N° 6/89, (Lei Assembleia Da Republica)", adopted on 15.4.1989.

Data set D1 for passenger information was not provided by Portuguese ports as none of them exceeded the threshold of 400 000 passengers per year.

Portugal was one of the first countries to fully comply with the Directive in respect of data collection and complete transmission. All required quarterly and annual data sets for reference years 1997 to 1999 have been transmitted to the Commission. The timeliness of data transmission improved over the years, and data for the first quarter of 2000 was received in line with the Directive. Problems with some codes that occurred in the beginning were solved.

9.11. Finland

Finland traditionally collects statistical information on maritime transport. The legal basis for this collection is a law of 1990 [15], amended in 1997 [16] for the purpose of the EC Directive. On 16 December 1996, the Finnish Maritime Administration adopted an administrative measure [17] for implementing the directive in the Finish statistical system.

[15] "Merenkulkulaitoksesta annettu laki (13/1990)", in force since 1.3.1990.

[16] "Laki merenkulkulaitoksesta annetun lain 1 ja 2 :n muuttamisesta (N:o 1248/1997)", in force since 1.1.1998.

[17] Published in the Official Publication of FMA 4 of 1.1.1997, in force since 1.1.1997.

Finland only needed a partial derogation for one variable (type of cargo) in 1997, and was the first country to achieve timely data transmission according to the requirements of the Directive.

All Finnish data sets for the reference years 1997 to 1999 and for the first quarter of 2000 were transmitted to the Commission. However, the data did not cover the Finnish domestic traffic, which should also be included according to the Directive. This information should be added in the future.

9.12. Sweden

In Sweden, the collection of maritime transport statistics is based on a law and a regulation on official statistics of 1992. [18] With a regulation on transport of persons and goods by sea of 1996 [19], and the establishment of a new survey in 1996, Sweden fully implemented the Council Directive in its national legal system.

[18] "Lag om den officiella statistiken" (SFS 1992:889), adopted on 4.6.1992, in force since 1.1.1993, and "Förordning om den officiella statistiken" (SFS 1992:1668), adopted on 17.12.1992.

[19] "Föreskrifter om uppgifter till statistik om person- och godstransporter inom sjöfart mm" (SIKA-FS 1996:01), adopted on 23 April 1996, in force since 1.5.1996.

For 1997, Sweden was granted total derogation for "number of units", and partial derogation for "nationality of vessel", "port of loading", "relation" and "type of cargo" (in data sets A1 and C1). For 1998 and 1999, it was exempted partially from providing "type of cargo" (in data sets A1 and C1) and totally from "units without load".

Sweden has transmitted all required quarterly and annual data sets for the reference years 1997 to 1999, and for the first quarter of 2000. The periodicity for data transmission foreseen by the Directive has not yet been met. Problems with conflicting codes in the first transmission batches were solved by re-transmission of all data sets.

9.13. United Kingdom

In the United Kingdom, the legal basis for collecting the data is the Statutory Instrument [20] passed in 1997. The collection of passenger data does not present a problem, as there are already existing systems for collecting them. For the freight data however, the UK had only run an annual, less detailed enquiry. Starting in 1997 the UK has designed a new collection system for the quarterly and more detailed data. The majority of data is collected from shipping companies and shipping agents, and some data are obtained from ports for statistical control purposes. Data providers may choose between five means of reporting, ranging from paper to full EDI methods. Around 90 per cent of all returns are submitted using electronic methods.

[20] "The Statistical Returns (Carriage of Goods and Passengers by Sea) Regulations 1997" (Statutory Instrument 1997 No. 2330), made on 24.9.1997, in force since 1.11.1997.

The UK was granted extensive derogation for freight data, i.e. total derogation for data set F1/F2 and derogation for information on port of loading/unloading, quarter and nationality of registration for all freight transport, and partial derogation for type of cargo.

The UK has already provided quarterly and annual passenger data for the reference years 1997 and 1998 and freight data with unknown destination for 1997 and 1998 to the Commission. Only data set D1 was supplied for all quarters of 1999. All data sets for the first quarter of 2000 were transmitted.

9.14. Norway

Norway is using the Statistics Act of 16 June 1989 No. 54 which gives Statistics Norway legal authority to collect the statistical information required by the Directive. No additional legislative measures were deemed necessary for applying the Directive. Statistics Norway planned to start reporting in 1999.

Norway has not yet transmitted any data to the Commission.

9.15. Iceland

Even though formal transposal of the Directive was not completed before winter 1998/99, Iceland was able to provide detailed information for 1998 in the annual data set A3, with information only on cargo and not on passengers. Reykjavik is the only Icelandic port for which detailed information is required.

Iceland also transmitted tables A1, A2 and C1 for 1998, but without relation, port of loading/unloading and gross weight.

In practice, the two main problems encountered in data collection in Iceland are obtaining information on individual vessels and data on destination and origin of vessels.

10. Conclusion

The experiences from the implementation of the maritime Directive have in general been positive. Setting up the reporting system in the thirteen EU Member States concerned, with a total of almost 1100 statistical ports, involved resources and efforts at all levels: the ports, the competent national authorities and the Commission.

The three-year transitional period turned out to be appropriate to the needs for adaptation. By January 2000, the majority of Member States had adopted the necessary legal and administrative measures and were ready for producing the required statistical returns. As at November 2000, the Commission has received the majority of data required by the Directive for the first three reference years. However, the collected information is incomplete as far as three Member States with important maritime transport are concerned. This has a negative impact on building EU totals and on dissemination of the data in general. The timeliness of data transmission still needs to be improved for guaranteeing a reliable and in-time dissemination of maritime data. Technically, the transmission of data to the Commission was continuously improved during the first three years, thanks to electronic transmission tools.

Given the provisional threshold for detailed reporting duty and the derogations granted during the transitional period, complete and comparable data on transport by passengers and cargo by sea will be available only from reference year 2000 onwards.

In a few countries there are still problems with non-response of data providers.

Compared with the time before the Directive, when only some of the Member States provided some general data on a voluntary and non-harmonised basis, the Directive constitutes an enormous advantage: for the first time, there will be regular information on maritime transport in the EU/EEA on the basis of harmonised definitions and classifications, and with a broad coverage of ports and indicators.

So far, the procedures foreseen in the Directive for laying down its implementing rules and for adapting it to economic developments have proven satisfactory. No changes in the text of the Directive are sought in short term. In the medium-term however, the pilot studies carried out pursuant to this Directive may lead to a need for changes. This will be explored once that more experience is gained with collecting the variables of the current system.