Restrictions for Free of Charge Dissemination Seen by Member States

Restrictions for Free of Charge Dissemination Seen by Member States

Doc. Eurostat/A-5/DWG/April04

DISSEMINATION WORKING GROUP

28th and 29th of April 2004

Bech Building, 5, rue Alphonse Weicker, Luxembourg-Kirchberg

Room Ampère

Beginning 10.00 a.m.

DWG/April04
DOC 11

Limits on free of charge dissemination

seen by Member states

Chairman :Gunter SCHÄFER

Page 1

Limits on free of charge dissemination seen by Member States

Country / Restriction / Reason / Date / Comments
Domain/Regulation
BE / No restrictions
/ Our guiding principle is that we only charge for overhead costs of producing publications, CD-ROMs, special data base extractions, extra statistical operations etc. and for a commercial use of our data. / Message (1/4/2004) / Question on commercial reuse of data in response message.
CZ / No restrictions known until now / Message (16/4/2004) / According to its dissemination strategy CZSO provides to public only aggregate information. We offer standard (i.e. prepared - publications, news releases) outputs free of charge via Internet; users have to pay when they need information in printed form or on electronic media (they pay cost for copying, printing, paper and so on). Fully paid are information prepared on demand - price is set by the time we spent on selecting data, analyzing* printing, copying etc. - for setting price we use Price List.
Just small remark to be clear - when user use internet site for finding information it will be free of charge, when our office does some selections from internet site, we will charge the user according our Price List.
DK / No general restrictions except very detailed urban statistics / Message (22/4/2004) / In general Statistics Denmark has no problem with your disseminating free data through "new-New Cronos". We already disseminate free data our selves to the geographical level of municipality.
However, I have not got reaction yet from the unit that produces urban statistics ( more detailed than municipality), so we might have some restrictions here. But I assume that this will not be in New Cronos either?
DE / General agreement to free dissemination but not all production units have given feedback on specific restrictions.
Restrictions known concern regional statistics. There needs to be a decision of the Working Group on Regional Statistics / Message (16/4/2004) / Ich habe zwar noch nicht alle internen Rückläufe aus den Fachabteilungen
bezüglich Ihrer Umfrage, aber es zeichnet sich ab, dass wir der free dissemination der an Eurostat gemeldeten nationalen Daten zustimmen.
Eine endgültige Aussage werden wir wegen der noch fehlenden Rückläufe erst an der Arbeitsgruppensitzung treffen können.
Eine Ausnahme bilden natürlich die Mikrodaten (ist wohl unstrittig),aber auch die Regionalstatistiken. Hier möchten wir erst die
Entscheidung der zuständigen Arbeitsgruppe "Regionalstatistik" bei Eurostat abwarten. Bis dahin lehnen wir für diesen Bereich die
kostenfreie Verbreitung durch Eurostat ab.
EE / No principle restriction / Message (22/4/2004) / There are no principal restrictions on the free of charge access to Eurostat's databases according to the dissemination strategy of the Statistical Office of Estonia. All statistics in the public database available on the website of the Statistical Office are free of charge. The same statistics could be published free of charge in Eurostat's databases.
Due to the small size of Estonia as a country the statistical confidentiality rule might appear more often than in bigger countries. Usually the data are labelled with "special flags" when sent to Eurostat.
We do not support free of charge dissemination of grid based and small area statistics compiled on demand of customers. Also all statistics where natural or legal person could be identified should not be available.
In the future, it could be a good idea to agree about the publication rules while planning a new survey or a project. The item might be an obligatory part of the legislation or agreement, even a gentleman agreement.
EL
ES / No restrictions
(The only potential restriction refers to microdata. Eurostat should respect INE's policy while selling INE's microdata in prices or terms of use) / Message (29/3/2004)
FR / No restrictions
(The only potential restrictions only refer to registers like SIRENE which are not included in Eurostat databases) / The reason for charging for SIRENE is that it is not a statistical product, and it is targeted to a very limited number of users / Message (23/04/2004) / - très fort développement de l'offre gratuite sur le site internet depuis 2 ans,
- cette offre se substitue en premier lieu à l'offre standard sur étagère, dans un second lieu à l'offre normalisée (qui est aussi standardisée dans son contenu, mais non disponible sur étagère), puis, en dernier lieu, à l'offre sur mesure (produit unique fabriqué à la demande du client),
- cette offre gratuite cannibalise d'abord le standard et, de façon croissante, le normalisé,
- avec la mise en ligne du web données locales et de la bdm2, les effets de substitution seront encore plus forts avec le normalisé,
- pour ce qui est des PSMs [Produits Sur Mesure], les effets de substition sont importants également : l'élargissement et la diversification de l'offre de contenus standard réduit les besoins en sur mesure et modifie les modalités de réponse ; l'articulation plus fréquente du sur-mesure sur une offre plus souvent gratuite ou peu onéreuse amène les clients à davantage se questionner sur la justification du coût plus élevé des PSMs .

Par contre, les sous-produits statistiques des opérations statistiques n'intéressant qu'un nombre limité d'entités (et, en pratique, nécessitant des exploitations spécifiques ou la constitution de bases de données ad-hoc) doivent donner lieu à facturation de l'ensemble des coûts de diffusion (y compris amortissement des dépenses de constitution des bases de données) puisqu'il s'agit d'un service rendu à un petit nombre d'entités et qu'il ne serait pas logique d'en faire supporter le coût à l'ensemble de la collectivité via l'impôt.
Enfin, les sous-produits non statistiques (de type listes d'unités : Siren, permis de construire, ...), lorsqu'ils peuvent être diffusés, sont évidemment tarifés puisqu'il s'agit à la fois de produits qui n'entrent pas dans la finalité de la statistique (et donc dans ce qui a motivé son financement via l'impôt) et de services rendus à un nombre généralement très limité d'individus.
IE / One possible area of concern for CSO is the distribution of small area population statistics (i.e. DED-level or NUTS-5). It would probably be acceptable to have some basic population statistics available at small-area level, but not the full list of variables for each DED. / We make this restriction because, at present, we earn a lot of revenue from commercial re-use of Census SAPS files so we would not wish to see the full detail of these files distributed free, / Message (29/3/2004) / As far as I know theses statistics are not disseminated in New Cronos. Old data (91) are very partially disseminated in the context of SIRE project through the Data Shops. (CK)
IT / No restrictions
CY
LV / No restrictions / Message (16/4/2004) / Confidentiality rules must be observed
LT / Data that should not be published free of charge:
- pilot survey data
- if the relative standard error exceeds 9 per cent
- Prices for consumer goods and services for the European Comparison Programme
- Prices for capital goods for the European Comparison Programme
- Estimated producer price indices / Message (20/4/2004) / Information that may be published :
- aggregated data
- when indication "confidential " is missing
LU / No restrictions / All published data (except tailor -made) is free of charge. / Message (21/04/2004) / Confidentiality rules must be observed
HU
MT
NL / No general restrictions / Message (16/4/2004) / All statistical information in StatLine and on the Website of Statistics Netherlands is free of charge. There are no restrictions on publication of those data by Eurostat. However, Statistics Netherlands supplies Eurostat with data that are not published on the website of Statistics Netherlands (and that are not included in StatLine). Those data deliveries are agreed on in Working Parties or laid down in Gentlemen Agreements. These agreements also stipulate which data are confidential. In general, the following types of data are not fit for free dissemination:
- Micro data. Micro data are a separate subject that is not part of this answer.
- Detailed data that are used by Eurostat to produce aggregates, e.g. some international trade data. The data deliveries of this kind of information are 'flagged'.
In general, Statistics Netherlands uses for its data deliveries to Eurostat the agreed formats and software. Ediflow and Stadium give information about the confidentiality for each data flow. If Gesmes can be used, it is feasible to give the status of confidentiality for each variable in a data flow. Of course, all this information is known to you. But we think it useful to emphasize the following.
- There are other Dutch institutes that supply data to Eurostat, for instance the Dutch Central Bank The deadline is too short to obtain all information of those institutes.
- We're more interested in a good working infrastructure than in the current picture. Eurostat has the tools (Ediflow, Gesmes, Stadium) to create an infrastructure that is dynamic and in which each institute can manage the confidentiality of its own statistical information.
- We can agree that the information in Ediflow, Gesmes and Stadium is used within the dissemination unit of Eurostat. We would appreciate it if Eurostat could inform us how the information flows in Eurostat are being organised (if possible, in the coming meeting of the Dissemination Working Group).
AT / Need to analyse the national legal situation. The Austrian statistical law foresees sales of statistics. / Da wir -wie Sie inzwischen wissen- ein eigenes Bundesstatistikgesetz haben, das von Eurostats derzeitiger Veröffentlichungspolitik abweicht, wir aber prinzipiell den allgemeinen Trend nicht durch aufwändigen bürokratischen Verwaltungsaufwand blockieren wollen, muss nun von unserer Seite her geprüft werden, inwiefern Eurostats Politik mit nationalem Recht vereinbar ist.
PL / General approval of approach. Reservations made for statistics exceeding the ‘Programme of Statistical Surveys of Official Statistics’. Separate consideration of commercial databases. / Message (23/4/2004) / Central Statistical Office of Poland (CSO) approves general outlines of Eurostat’s proposal to prepare free-of-charge database containing statistical information about EU countries.
We find it essential to exclude both the data covered by Polish statistical confidentiality regulations and the data exceeding the Programme of Statistical Surveys of Official Statistics.
The issue of free access to data covered by our commercial databases requires separate settlement.
CSO would support Eurostat with free of charge data available on our website and in publications (electronic and paper versions). We request Eurostat’s cooperation on mutual terms.
CSO will take the official position on receiving detailed list of data required by Eurostat.
PT / The potential reservations to free dissemination concerns only the information about economic activities and international trade were the information available by Eurostat is more detailed that the one free of charge on INE-P;
So, ours restrictions concerns mainly the dissemination of a lower level of aggregation, i.e., more than 3 digit level on NACE and 2 digit level in NC; / Message (19/4/2004) / The majority of data published by Eurostat is disseminated on-line at a national level and aggregated by NACE. INE-P assures the same dissemination policy on free services;
In INE-P, we put free of charge, in our web site, the data at 3 digit level on NACE/CAE digit level on NC and lower levels are paid by the users.
SI / No restrictions detected until now / We don't have any reservations considering free of charge access. For our data, marked as confidential, the disclosure control (hiding cells) should be done in databases as it is done already. My colleagues from subject-matter fields have checked some important data in New Cronos and there are no major problems. Still, it was not a detailed check so maybe some things can be discovered later and I will let you know if there will be something important. / Message (16/4/2004)
SK / General agreement to free dissemination.
Restrictions are seen in free dissemination of entire Yearly National Accounts, Input/Output Tables and of all statistical information in more detailed structure as it is normally published;
moreover for deeper classifying than 3 digit NACE (eventually by the CPA) and 2 digit CN of foreign trade. / SO SR is obliged by the budget rules of the SR to gain incomes from statistical data dissemination and to transfer them to the State Budget. / Message (23/4/2004) / Free statistical data dissemination from government department statistical surveys was not discussed up to now.
FI / General agreement with free of charge dissemination.
Limits are seen in:
- If there is an unfinished project where dissemination and prices have already been discussed, the agreed prices and dissemination should be upheld. An example of this is the Urban Audit project which contains data that are not free of charge in all countries. Urban Audit data are partly chargeable at Statistics Finland. Agreements should also always contain decisions about dissemination already at the data collection stage.
- The lowest threshold limit for small area data (including data from small municipalities) is 150,000 persons according to the NUTS regulation. This is mainly governed by data protection requirements.
- Data that are not in the format of statistics, such as digital geographic boundary data and classifications have to be discussed separately. - Digital geographic boundary data cannot be disseminated without a separate contract and free of charge. Basic classifications are not charged for.
- Statistical grid square data and subarea data cannot be disseminated free of charge. These datasets might also contain confidential data. / Message (16/4/2004) / Other remarks:
- Statistics that include “flagged” information, such as confidential, uncertain or estimated data, should be disseminated only at the aggregate level. In other cases there should be appraisals of the level at which information should be supplied to Eurostat. More detailed information is required from Eurostat concerning the supplying and using of “flagged” data.
- Microdata for research purposes should be chargeable (Commission Regulation EY 831/2002).
- Data that are confidential by virtue of legislation should not be disseminated at all. For example, trade statistics (Intrastat) and information where an individual person or enterprise could be identified should not be made available at all. Examples of confidential statistics are those on innovations and on research and development.
- It is unlikely that statistics that are delivered to Eurostat by other governmental data providers, as well as those received by Statistics Finland from the enterprise sector under separate contracts could be disseminated freely. There has to be enough time to check the reservations of other national statistics providers. There should be a common way of defining each variable separately.
SE / Restrictions with respect to level of detail for trade data. SE would not like to that Eurostat distributes trade data with higher level of detail than 8 digits.
Concerns about confidentiality mentioned under other remarks do not really affect the free of charge dissemination. Micro data is anyhow excluded from it and confidentiality must be observed as before. / Statistics Sweden considers the CN 8 digit data in Intrastat to be too detailed since the quality of the data is often poor on detailed level. Therefore, we have decided only to publish our official statistics on external trade on six digit level. We would prefer the EU-statistics on Intrastat also to be on that level. The access to data free of charge should only be given to data on 6 digit or even more aggregated level. Statistics Sweden is not in favour of the proposal of Eurostat to provide detailed data on 8 digit level free of charge on Internet. / Message (15/4/2004) / Further remarks:
Goods on road:
Statistics Sweden reports microdata on transport of goods on road to Eurostat according to EU Regulation 1172/98. These data should undergo confidentiality checks, data can not be disseminated if the number of observations is below 10. These rules are also stated in regulations.
Maritime transport, national and international:
Before disseminating these data, common statistical confidentiality rules should be applied.
UK
CH / Issues seen for:
- Geographic data
- Reuse of data for commercial purposes / Statistics Switzerland has a system of distribution on the basis of sales, however, the data delivered to Eurostat is not concerned except the two points mentioned in the previous column. / Message (16/4/2004)
NO / No restrictions / Message (15/4/2004)
BG
RO / No restrictions / Message (16/4/2004)
TR / No restrictions / Message (16/4/2004) / Moreover, micro data including personal information are not being distributed. After eliminating the personal information, it shall be supplied to the users with the approval of the President of the SIS.
On the other hand, harmonisation studies are being carrying out in the context of the project titled" Upgrading Statistical System of Turkey". In addition we are also filling out and sending back Eurostat's questionnaires on many topics, regularly. However by examining the database it is realised that not all the data we are sending are included on Eurostat's web site.
In this context, we can cover all data you received from our Institute for the dissemination from Eurostat's web site.

Page 1