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. Description of changes is found under 'Comparability - over time'
Having an almost complete Register of Buildings and Dwellings is rather unique in an international context. Census of Housings is internationally most often compared with Censuses of dwellings based on questionnaires. In these comparisons, we have problems identifying institutional dwellings.
The traditional population and housing censuses included only occupied dwellings. With the possibilities to integrate administrative registers we are now also able to obtain figures on non occupied dwellings. From 2010 also non occupied cottages are included.
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:
From 1981-1987; the type of ownership was not maintained by linking data with the Real Property Taxation Register (ESR), but only due to the municipal administration of building and development cases. 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 the other ownership categories rises. In 2012 and 2013 there was a problem obtaining correct ownership on dwellings especially those owned by non-profit building societies and housing societies. This was caused by the Land Registration being digitized. These errors are corrected in the 2014-data. There is still 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 has been 18. Since 2007 youngsters under 25 years living with their parents/parent are counted as part of their family.
At the beginning of 1980, the municipal administration made frequent changes to the road codes. This caused some problems, following the same dwelling address back to 1981 and resulted, to some extent, in double counting. A changed road code made it look like a migration of the households though this of course was not the case.
Until 2005 dwellings were divided into to groups; 'proper dwellings' and 'other dwellings'. Proper dwellings are dwellings intended for permanent residence. Other dwellings consists of cottages and residential buildings for communities. After 2005 a dwelling is defined as part of one or more buildings, which has a separate address and is intended for permanent residence or is used as residence for at least one person according to CPR at the time of reference. Cottages which are occupied at the reference time is counted as occupied dwelling, while unoccupied cottages are not part of the census between 2005 and 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.
From 2010 the input data from BBR has another architecture and a new way of matching addresses between BBR and the CPR is introduced to reduce the number of households without dwellings. Data for the year 2010 is provided in both series for comparison.
In summary the most important changes are:
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:
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)