Comparability

Comparability - geographical

Denmark bases its population statistics on an administrative register, which also applies for a few other countries, whereas many countries take censuses every five or ten years and make estimates in the interim period. The statistics are register-based and data are released 30 days after the end of the census period to include late reported events. Other countries with register-based censuses may use a shorter/longer delay/time delimitation period. The longer this period is, the more accurate the statistics will be. Thus, the time delimitation period is a choice between current and timely statistics on the one hand and accuracy on the other. Eurostat publishes statistics for Denmark on immigration and emigration in compliance with Regulation (EC) No 862/2007. These figures are lower, since other time criteria are applied regarding the duration of the stay. Only people who intend to stay in Denmark for 12 months must be registered as immigrated in Eurostat’s statistics. In Denmark’s national releases, this is three/six months for immigration and six months for emigration based on the rules for registration in CPR.

Comparability over time

The population figure for the period 1971 and onwards is based on the same source, i.e. the Civil Registration System (CPR). The population figure from before this period is based on censuses. The population figure for Denmark is comparable from 1769 and onwards. The local government reform of 2007 in Denmark reduced the number of municipalities from 271 to 98. For 19 of these new municipalities, it is not possible to make an exact comparison before and after 2007. Approximately 20 per cent of the population lives in these 19 municipalities, which are composed of several divided municipalities from before the local government reform. From 2007, delayed reporting is no longer included in the current quarter. A survey of the changes of address for the year 2007 has shown that 2.5 per cent of the year’s reported changes of address can be attributed to previous years but were not registered until the current year. Reporting of events to CPR and the civil registers has subsequently become timelier than at the beginning in 1971. In Q4 1992, Statistics Denmark reduced the period of waiting for delayed reporting from 40 to 30 days. Reducing the time delimitation has had the biggest impact on emigration, which accounts for the most delayed reporting of events. From June 2016, immigration and emigration is recorded according to when the migration was registered. This means that all migrations are included in the statistics regardless if the migration has happened earlier. With the old method applied from 2007 to the beginning of 2016, the late registrations meant that these emigrations were not included in the emigrations for the quarter/year, nor in subsequent assessments. For that reason, emigration has been underestimated with up to 33 per cent in the quarterly statistics and up to 17 per cent in the yearly statistics. In the population accounts, these late emigrations in 2007-2016 have been partly saved by adjusting the population in the tables at the beginning of the year for these late migrations along with other adjustments regarding previous years. In practice, the population at the end of the year has not equalled the population at the beginning of the following year. From June 2016, the population at the beginning of the year equals the population at the end of the previous year. Special conditions apply for the third and fourth quarter of 2019 where emigrations among foreign citizens are strongly influenced by the fact that the former minister of economic affairs and the interior had asked the municipalities to check the residence registrations in CPR focusing especially on foreign students and employees. For more details, read the documentation of statistics for 2019.

Causes of death before and after 1994 cannot be compared directly. This is because the cause of death codes have been changed. Up to and including 1993, ICD8 (International Classification of Causes of Death 8. rev) was used. As from 1994, IDC10 is used. From 15 June 2012, it has been possible for two people of the same sex to get married and subsequently divorced. At the same time, the possibility for two people of the same sex to enter into a registered partnership was no longer available. This also has an impact on the statistics of dissolved partnerships and same-sex divorces.
The family definition was changed in 2008. The family statistics with the new definitions exist back to 1986. It is not comparable with family statistics produced and published before 2008. The change consists in family statistics produced before 2008 using a family definition where children included in the families of the parents had to be under 18 years, whereas now it is 25 years. This change increased the number of children living at home and reduced the number of singles. In 2008, there were 164,000 children living at home at the age of 18-24 years. The change does not affect the number of households, but has some impact on the distribution of households on types of households. The development in the number of changes to Danish citizenship is influenced by e.g. the rules regarding immigration. When a large group of immigrants comply with the duration of stay requirements to obtain Danish citizenship at practically the same time, this will often show in the number of changes to Danish citizenship. For most people changing to Danish citizenship, Danish citizenship can only be acquired through adoption of an Act of Notification of Naturalization. The Danish Parliament usually adopts two such bills each year. However, bills are sometimes postponed. For example, this was the case in 2007, where the general election on 13 November 2007 meant that only one bill on naturalization was adopted in 2007, but three in 2008. Thus, the parliamentary work is also a factor that decides if specific changes in citizenship are recorded one year or the other. In January 2014, a changed calculation method of fertility rates was brought into use. The change was made to a cohort-based approach where age of mothers is calculated by the end of the year instead of at the time they give birth.

Coherence - cross domain

Under National Church, statistics are also available on the population figure and the development of the population by parishes, which can be aggregated to deanery and diocese. Urban Areas focus on population figures and population density broken down by cities of various sizes and rural districts. The cities may transcend municipal boundaries. The population statistics are used to prepare population projections broken down by municipalities, ancestry etc. The municipalities prepare their own statistics based on information from their own local civil register. These statistics do not always correspond fully with those of Statistics Denmark, since the municipalities do not wait 30 days for delayed reporting of births, deaths, changes of address and migrations etc. The differences between the statistics from Statistics Denmark’s and those of the municipalities can thus be explained by the differences in delimitation and extraction time in relation to reference time. The household concept in the statistics of households and families deviates from the household concept in the housing stock, which is based on the concept of dwelling household, and from the concept of housekeeping unit applied in the Household budget survey, which is an interview-based survey involving a limited number of families. The housekeeping unit concept is closest to the UN recommendations, but requires questionnaires or similar and is therefore not compatible with all-encompassing register-based statistics.

Coherence - internal

The sum of the quarterly figures for population and events is not consistent with the year. This is because the annual statistics are assessed at a later point in time than Q1, Q2 and Q3. As a result, the annual statistics are more up-to-date regarding late reporting. The population accounts are available in Statbank Denmark table BEV22. The population accounts show the correlation between the population at the beginning and end of the quarter and the intermediate figures for live births, deaths, immigrations and emigrations. The adjustment item can be fully accounted for by subsequent municipal registration as well as missing and found persons.