Skip to content


Contact info

Department of Social Statistics, Population and education
Annika Klintefelt
+45 39 17 36 78

Get as PDF

Census of Housing

Then Census of Housing has been based on the BBR and CPR-registers since 1. January 1981. With the exceptions of a few variables the data are consistent from 1981. There are though certain reservations for comparison during this time period.

Comparability - geographical

Census of Housings is internationally most often compared with Censuses of dwellings based on questionnaires. In these comparisons, we have problems identifying institutional dwellings.

Comparability over time

The questionnaire based population and dwelling census included only the occupied dwellings (households), while the current statistics also contain information on non-occupied dwellings. The dwelling statistics has been calculated based on BBR and CPR dating back to January 1, 1981. Many variables have had the same linguistic content throughout the period, but there are certain reservations for comparisons over the time:

From 1. January 1988 a monthly update of the ownership via the ESR was performed and later the update has been performed three times a year of which the last one is performed in week 51.

From 2010 also ownership of owner-occupied flats is available (see definition), which means that the number of dwellings with all ownership categories rises. In 2012 and 2013 there was a problem obtaining correct ownership on dwellings. These errors are corrected in the 2014-data. There are still however, too few dwellings owned by non-profit building societies in some municipalities. In 2010-2015, far too many dwellings owned by non-profit building societies where registered in the Municipality of Frederiksberg.

Until 1992, persons under 26 years living with their parents, were counted as children. From 1992 to 2006 this age limit was 18. Since 2007 youngsters under 25 years living with their parents/parent are counted as part of the parents family. In the early 1980s, there was a significant change in the road codes in the municipalities, which led to a problem of following the same physical address all the way back to 1981. A changed road code would thus be perceived as a move of the person, although this has not physically been done. This could in some cases result in a double count of dwellings.

Prior to 2005, housing was generally divided into "proper dwellings " and "other dwellings". Proper dwellings are dwellings intended for year-round housing. Other dwellings consists of holiday homes and residential buildings for communities. As of 2005, a dwelling is defined as part of one or more buildings for which there is an independent address and which is intended for or used as a housing at the reference time. Holiday homes inhabited at the time of reference are included as housing, but holiday homes without CPR registration are not included in the housing statistics in the years 2005-2009 / 2010.

In the period 2005-2010 the BBR-amendment register was used for collecting information on not yet completed dwellings, where persons according to the CPR already had their address.

The reporting and structure of BBR data changed during 2010. In addition, efforts have been made to improve the match between BBR and CPR addresses in order to reduce the number of people without housing. Data for 2010 is calculated using both 'old' and new method. Summarized here are the main changes:

  • Number of unoccupied dwellings is reduced by almost 23.000 dwellings of which 3.900 were unoccupied residential buildings for communities that are not considered as dwellings.
  • Data for cottages not used for permanent living is now included.
  • More people are linked to dwellings. This is partly because there, in some dwellings, are placed more than one household.
  • Ownership of owner-occupied flats are now available. This category is not part of the ownership variable anymore and the numbers in the other categories has therefore increased.

In 2017, BBR was rearranged to a completely new register. In this connection, significant implementation problems have emerged. A completely new data structure has been introduced, which Statistics Denmark has not received a total documentation of. Further, a new set of codes has been introduced. It has not been possible to generate 2017 figures based on the new BBR reports. Therefore, the new method/new reporting cannot be distinguished from the underlying trend in the housing stock between 2017 and 2018. In connection with this, the number of people who cannot be matched to a home has risen from 28 000 to 33 000. One reason seems to be being that the addresses in the CPR are not always updated.

Coherence - cross domain

Because the housing census uses the Central Population Register as a data source, there is a nearly complete consistency with statistics on population and statistics on households.

However, the consistency with the statistic on constructions is, for several reasons, somewhat lower:

  • The housing census counts the number of dwellings on January 1, indicating the construction year of the dwellings. The difference in the number of dwellings between two years is a net figure (registration and deregistration). The statistics on constructions give a gross figure for the number of new dwellings.
  • Because of the municipal administration, it is not possible to make figures for combined, divided and demolished dwellings.
  • Changes in information, e.g. between occupations for residence or non-residence is undertaken directly in that part of the Central Register of Buildings and Dwellings keeping the existing buildings and dwellings.-
  • One major problem involved in compiling the statistics on constructions concerns delays.
  • The definition of a dwelling differs between the two statistics. In the construction statistics a dwelling is a unit reported as a dwelling by the administrative authorities. In the dwelling statistics a dwelling can also be a unit not reported as a dwelling by the administrative authorities. In this case, the criterion is that the address of a unit matches an address of a household.

Coherence - internal

Inconsistencies between tenure, ownership and the result of the match between BBR- and CPR-addresses (dwellings with registered population) are corrected. Less than 1 pct. of the dwellings get another tenure (to either unknown or unoccupied)