Note prepared by the secretariat
2. The Provisional Agenda was adopted.
3. The following substantive topics were discussed at the meeting:
4. Mr. W. de Vries (the Netherlands) was elected as Chair, Mr. D. Meuldijk (the Netherlands) as Vice-Chair. Mr. F. Thomas (United Kingdom), Mr. J. Byfuglien (Eurostat), Mr. J. Sobel (United States), Mr. V. Schreurs (Eurostat) and Mr. T. Davis (Canada) were discussants for topics a) - e).
5. The topics were discussed on the basis of the papers prepared by Canada (4), Finland (2), Hungary, the Netherlands (5), Italy, Poland, Romania, Russian Federation, Sweden (2), United States (2), United Kingdom, Eurostat (2) and UNSTAT.
6. The secretariat distributed the portion of the integrated presentation of international statistical work programmes in the ECE region relating to the project on geography in the Conference's programme of work (CES/830/Add.2), and drew the attention of the participants to the goals that the Conference has identified for work in this field.
7. The Work Session recommended that the following be included in the 1995/96 programme of work of the Conference of European Statisticians:
2.6 GEOGRAPHY
Activities and means:
Activities of ECE: (CES Priority A)
8. The Work Session recommended organizing the next Work Session in March 1996. The US Bureau of the Census was asked by the participants to investigate whether they would be able to host the next meeting. It was further agreed that the continuation of the work including the next meeting will be prepared by an organizational committee composed of representatives of Statistics Canada, Statistics Finland, Statistics Netherlands, United Kingdom, US Bureau of the Census, Eurostat and the ECE secretariat.
9. The other conclusions reached at the Work Session on the substantive items of the agenda are summarized (in English only) in the Annex.
Documentation: Reports prepared by the Netherlands, Sweden, ISO and the United Kingdom.
1. The Work Session expressed the opinion that it would be useful to make available information about software which can overcome, wholly or partially, problems due to non-standard geographic information. The countries were requested to provide details to the UN/ECE secretariat, who will compile a list of such software.
2. The Work Session noted with interest the information on the activity going on in the field of international standardization of geographic information -- especially in the Technical Committee (CEN TC) 287 of the ISO-European Committee for Standardization (CEN) and the preparatory work for the ISO Technical Committee (ISO TC) 211. Participants stressed that it would be important if the representative of CEN TC 211 could inform future Work Sessions on the outcome of the CEN TC 211. The UN/ECE secretariat was given a "Liaison A" status with the ISO TC 211, and it will provide countries with the documentation on ISO TC 211 when required. The opinion was expressed that it would be highly desirable if the chairman of the ISO TC 211 could participate regularly in future Work Sessions. Furthermore, the view was expressed that the countries which will take part in the work of the ISO TC 211 should report on how their efforts have influenced its activities at future Work Sessions.
3. The Work Session recommended that those countries with relatively advanced statistical applications of GIS would investigate the possibilities of whether they could take part in the work of ISO TC 211. The UN/ECE secretariat was requested to approach the Heads of the National Statistical Offices of Canada, Finland, Sweden, Norway, the Netherlands, United Kingdom and United States inviting them to nominate national representative(s) in individual working groups of the ISO TC 211. The choice of working group(s) will be determined by the expertise and experience of national representative(s).
4. The Work Session noted that special funding will be necessary for this work. The UN/ECE secretariat was requested to investigate whether Eurostat would be willing to financially support these activities.
5. There was a common opinion that the next Work Session should consider issues of statistical interest for the ISO TC 211. The list of items formulated at the Ottawa 1993 meeting would be used as a basis (see CES/795 Annex II).
Documentation: Reports prepared by the Netherlands (3), Poland, Romania, Russian Federation, Sweden and Eurostat (3)
6. The Work Session attempted to formulate a definition of statistical and geographical functionalities. While statistical functionality focuses prevailingly on the use of statistical functions and operations, geographical functionality deals with spatial aspects of data.
7. The Work Session also discussed for which phases of statistical production the use of GIS could prove to be useful. In the data collection phase, the pos- sible designing of sampling frames, definition of data collection units and data validation were mentioned. In some applications GIS was also successfully imple- mented for different statistical calculations and for data analysis. With regard to application areas, particularly censuses, land use statistics, environment and human settlements statistics were mentioned.
8. The most important use of GIS, however, is in the linking of different data sets using a spatial reference, and the dissemination of statistics to different users. Some participants expressed the view that users should learn to understand output maps more as a form of statistical tables.
9. The Work Session also noted some problems of integration. From the methodological point of view, especially different data models, different time handling and diverse aggregation methods in statistics and geography were men- tioned. The problem of the comparability of geographical entities was also linked to them.
10. The Work Session pointed out some important georeferencing methods used for GIS in statistics such as polygons, grid cells, point coordinates, census units and administrative units, where the latter is widely used. However, some participants mentioned that referencing to administrative units does not have a stable character.
Documentation: Reports prepared by Canada, Finland, the Netherlands(2), United States (5), Italy and Eurostat.
11. The Work Session noted that approaches to define rural and urban areas in individual countries very often differ. Some participants expressed the opinion that a simple distinction between rural and urban areas is no longer relevant. Some views were also expressed that in many countries, the concept of urban areas was widely used and politically important.
12. It was mentioned that specifications of geographic areas are also dependent on what authorities are responsible for their definitions, and what criteria are used in this context. Furthermore, the definition of areas could be influenced by such factors as the nature of data and their level of accuracy.
13. The Work Session also noted that the diversity of national approaches can cause difficulties in attempts to integrate these methodologies at the international level. The open question remains: how far is international standardization desirable and possible?
14. Data confidentiality, methodological issues and data availability are other important problems to be considered when defining concepts of geographic areas. Particular attention was drawn to the questions of data confidentiality, especially when small area information from different data sets is linked together. There was common agreement that this issue should be carefully studied in the future.
Documentation: Reports prepared by Canada, Hungary, Sweden and UNSTAT.
15. The Work Session noted that while some applications were based on commercial software packages, in some cases tailor-made software packages are also used. Many participants expressed the view that the exchange of experiences in the use and evaluation of GIS software packages would be highly desirable.
16. The Work Session recommended that the Conference establish a special group of experts, the aim of which would be to further increase knowledge on GIS software tools and concepts suitable to statistics. The group would focus on technological as well as on methodological issues. To facilitate the work of the group it was recommended to use to the largest possible extent electronic means.
17. The representative of Canada expressed willingness to steer the work of the group. The Netherlands, Denmark, Sweden and Eurostat expressed their readiness to participate in this work. It was agreed that the group will continuously disseminate the information on the outcome of the joint work to other countries. Furthermore, the group should report to the next work session on its progress. The UN/ECE secretariat offered to assist in setting up and maintaining this network.
Documentation: Working papers prepared by Canada, Finland (2), United Kingdom and United States.
18. The Work Session noted that some specifics in marketing of geo-referenced information can be identified. For example, to the extent that GIS allows agencies to disseminate statistical data for very small areas, it is more important to ensure that data are not misused.
19. The presentations showed that approaches may differ. While in some countries the statistical agency itself takes over the initiative of selling the information (retail marketing), in other countries external vendors are involved (wholesale marketing).
20. The following reasons for the marketing of statistical information were mentioned: revenue, return of investments paid by taxpayers and direct feedback to statistical agencies.
21. There was general agreement that an effort should be made to avoid keeping prices very low. However, it is sometimes difficult to avoid underestimation of prices, since it is not always easy or appropriate to incorporate all costs of statistical agencies in their calculation.
22. The Work Session noted that two major subjects of marketing can be mentioned: basic data and final outputs of statistical production. While basic data, not having a value-added character, could have rather low prices, the outputs of statistical analysis should have a higher value and could therefore be of more interest for marketing.
23. Other issues considered during the discussion were: competition between private and public sector, short- and long- term profitability, motivation of statistical staff, advertising of statistical information, etc. It was concluded that since marketing of statistical information is still in its infancy, this issue should be studied further.
24. The representative of the General Register Office of Scotland, United Kingdom, informed the Work Session about the International Seminar on GIS held on 22 and 23 September 1994 in Edinburgh, United Kingdom. The Seminar was attended by eight countries and was devoted to the GIS implementations in censuses. The agenda of the meeting covered many topics of a character similar to those considered at the Work Session. The seminar was the fourth event of this kind (since 1991). It was agreed that to better integrate the work in this field there should be close cooperation in the preparation of meetings in future. Furthermore, it was proposed to organize the next Work Session back to back with this meeting of experts.
25. The representative of the US Bureau of the Census informed participants about the Urban and Regional Information Systems Association (URISA) dealing with GIS. He mentioned that the outcome of its activities may be of interest to the Work Session.
26. The Work Session expressed its gratitude to the Central Bureau of Statistics of the Netherlands for having hosted this meeting and for the excellent organization and working atmosphere. It also expressed its appreciation to discussants and to all authors of papers and demonstrations.