Sven Egmose

Statistics Denmark

Relations to the administrative business register

In Denmark the Statistical Business Register, SBR, is integrated with the administrative business register through an on line connection. All units and data within the administrative business register are also to be found in the SBR. On top of that the SBR contains more unit types and more data. The administrative business register contains legal units and production units, i.e. LKAUs belonging to the legal units. The production units were introduced in the administrative business register at the request of Statistics Denmark, and from the start of the administrative business register in 1999 it was not foreseen that the production units should be used intensively for administrative purposes.

The situation has changed. Now a range of administrative bodies use the production units for administrative purposes. And they demand a more complete coverage of activities on all addresses than what is now in the register. The administrative bodies see only one way forward to reach this, namely to open up for self registration of production units on the web and maybe even introduce an obligation for the legal units to register their production units.

This raises some problems, because the production unit is defined according to guidelines given by Statistics Denmark which among other things means that the rules of continuation are laid down according to statistical considerations. Continuation of production units is not exclusively connected to the owner. A production unit can remain the same (and keep the same id-number) even if it is taken over by a new owner. This is a very big problem if a self registration on the web is introduced, and the consequence is that the definition of the production unit probably will be changed. If this will be the case it will be a big problem for Statistics Denmark. In order to keep the SBR in the same good shape as it has now we will have to cut the very tight integration with the administrative business register and make a re-designing and re-programming of our SBR and it will take considerable resources.

Quality

Technically the SBR is now a well-functioning system. The inappropriateness which has been in the system is remedied, and the updating procedures are efficient. This means that the conditions for improving the quality of the content of the register are right, and efforts are undertaken to make use of that.

At the end of 2003 Statistics Denmark undertook an investigation of the quality of some of the features of the SBR. The results from this investigation were presented at the Eurostat Working Group meeting on SBR in June 2004 and the report can be found on the Eurostat home page. The main results were not impressive. At the most detailed level of activity codes some 19 percent of the activity codes for the enterprises were wrong. The wrong codes were more frequent among small enterprises than among the bigger ones. For enterprises with more than 10 full time employees some 8 percent of the activity codes were wrong.

This report is of course made available for our users within our institute, but besides that we have started a home page (for internal users within Statistics Denmark) on the state of the art of the SBR. Besides information on the actual status in the form of number of active units, this home page also informs on number of updating made within the latest month, and with a possibility to see which business units have been subject to i.e. a change in activity code within the latest month.

Response burden

The response burden posed on business units by the Statistical Institute has been a very hot issue during a number of years. And the issue has become even hotter during the last year.

Since 1998 Statistics Denmark has every year published figures on the response burden posed on business units by Statistics Denmark. These figures have been based on estimates on the average time used to fill in the different questionnaires. This year a new measurement on the burden has been introduced. The new measurement is based on interviews with a number of enterprises who were asked about the time used for a range of activities performed to fill in questionnaires for the different statistics. This measurement was made by a private company and the results were published this summer. The result of the new investigation was a somewhat higher response burden than the former investigations. The results were in no way dramatic and they underlined that the burden posed by the Statistical Institute is real modest compared to other administrative burdens even though the statistical burden is very highly profiled in the public discussions.

Even if the statistical burden on enterprises is rather modest a lot of efforts are made by Statistics Denmark to reduce the burden. One tool for that is to open up for electronic reporting. In Denmark a new portal for all communication between public bodies and the business units has been established. Statistics Denmark is one of the pioneers in this work. Unfortunately the portal has run into severe technical problems which mean that we are still waiting for electronic reports on a large scale to come from the business units. The expectations from the politicians and the organisations are that electronic reporting will reduce the burden substantially.