Housing Stock and Housing Tax in a Regional Perspective

Jan Overgaard Olesen and Erik Haller Pedersen, Economics

INTRODUCTION AND SUMMARY

The housing market is often the focus of public debate. This article draws a general picture of the Danish housing stock, housing wealth and taxation of real property. The article is based on special data extracts from Statistics Denmark with details of the living space represented by the Danish housing stock, as well as of the regionally-distributed revenue from taxation of real property. The data concerning living space was previously the basis for statistics for housing wealth in Denmark^[1].

The article focuses on regional aspects. Dwelling conditions are distributed geographically on municipality groups, outlining the development in living space and value of dwellings over time. The outline is distributed on types of dwelling, including owner-occupied and rented dwellings. Finally, a distribution of the property value taxation revenue on municipality groups is presented, and effective tax rates are calculated.

There are considerable variations in housing consumption in various areas of the country. On average, a resident of a rural municipality has 15 m² more living space available than a resident of Copenhagen or Frederiksberg. Moreover, there is considerable geographical variation in terms of types of dwelling since rented dwellings are mainly located in larger towns and cities, which is also where cooperative housing is most prevalent.

The effective taxation of the value of real property in terms of revenue as a ratio of owner-occupied housing wealth has been falling since the early 1990s and will this year reach an estimated level of 0.5 per cent. The decrease in effective taxation was greater in the 1990s than it has been under the tax freeze from 2002 and up to the present time. The tax freeze in its present form will, however, result in an ongoing decrease in the effective property value tax rate for as long as housing prices continue to rise.

Since the start of the 1990s the strong increase in housing prices in the Greater Copenhagen Region in its broadest sense, i.e. including the Counties of Frederiksborg and Roskilde, has led to a regional shift in the revenue from property value taxation, so that the Greater Copenhagen Region now accounts for a larger share. Nonetheless, the effective property value taxation has decreased most strongly over time in this particular region.

Housing stock
There are many different types of dwelling in Denmark, and there are considerable regional variations in the population's dwelling conditions. Data from Statistics Denmark provides for an overall description of these regional variations. In this article the data is divided into eight municipality groups as defined by the Danish Customs and Tax Administration (Skat), cf. the Appendix for an overview. Municipality group 1 comprises the City of Copenhagen and the City of Frederiksberg, while municipality group 8 consists solely of rural municipalities. In principle, the less densely built-up the municipality, the higher its municipality group number. This classification ensures a more homogenous distribution of the housing stock in terms of price setting than distribution solely on the basis of geographical criteria. See Box 1 for further details of the data material.

Around half of the Danish housing stock compiled in terms of living space (number of square metres) consists of owner-occupied detached one-family houses and terraced houses, cf. Chart 1. Since the living space in these houses on average exceeds that for other types of dwelling the share in terms of the number of dwellings is smaller. Rented dwellings, including social housing, account for almost 30 per cent of the housing stock, while the remainder is distributed on e.g. cooperative housing, freehold flats and holiday dwellings.

Most of the Danish housing stock is situated in small towns and rural municipalities (municipality group 8), cf. Chart 2. This is related to the fact that these municipalities cover the largest part of Denmark in terms of area, although the housing density is lower. The living space is relatively evenly distributed on the other local government groups. The distribution of the living space on types of dwelling and geography can be interpreted to show that the most prevalent dwelling in Denmark is a detached one-family house in a rural municipality.

Distribution of the housing stock on owner-occupied and rented dwellings and on municipality groups shows that rented dwellings (privately and publicly rented dwellings and social housing) are mainly situated in larger towns and cities, with a large concentration in the Greater Copenhagen Region, cf. Appendix Table 1. 38 per cent of the population reside in municipality groups 1, 2 and 4, but this area contains 46 per cent of the country's rented living space.

The largest proportion of freehold flats and cooperative housing is also found in the Greater Copenhagen Region. Together with privately rented dwellings, cooperative housing dominates in the City of Copenhagen and the City of Frederiksberg, which each accounts for approximately 30 per cent of the living space. Holiday dwellings make up a significant share of the housing stock in a number of rural municipalities, i.e. municipality group 8.

For some years the fastest growth has been seen in the stock of owner-occupied dwellings. Since 1981 the total owner-occupied living space in Denmark has increased by one third, while the total rented living space has increased by only half as much. In Copenhagen and Frederiksberg (municipality group 1) the dwellings built during the last 25 years have almost exclusively been owner-occupied, but nevertheless the latter still account for only approximately half of the total living space, cf. Chart 3. For comparison, owner-occupied dwellings account for 66 per cent of the total living space on a national basis.

The distribution between owner-occupied and rented dwellings is influenced especially by new construction, although other factors also play a role. For example, in connection with " parent purchases" dwellings will in principle change status from owner-occupied to rented, cf. Box 2.

Cooperative housing is here classified as owner-occupied dwellings since in economic terms it in several aspects most resembles owner-occupied dwellings. This applies for instance to mortgage-finance opportunities, as well as the opportunities and risks related to capital gains and losses on the dwellings. Cooperative housing is the category that has seen the largest increase over the last 25 years, by approximately 10 per cent per annum for the country taken as one, which is far more than for other types of dwelling, although the increase has been from a low level. The total living space has increased by only 1 per cent per annum, cf. Appendix Table 3. 

On interpreting the " owner-occupied housing rate" in Chart 3 it is important to note that this does not reflect the number of dwellings, but the number of square metres. Since owner-occupied dwellings are on average larger than rented dwellings an owner-occupied housing rate calculated in terms of number of dwellings instead of number of square metres will result in a lower ratio of owner-occupied dwellings to all dwellings. Compiled in terms of number of dwellings the owner-occupied housing rate (including cooperative housing) for Denmark overall was thus almost 60 per cent at the beginning of 2005.

Historically, migration has been from country to town, and during the last decade especially the Greater Copenhagen Region has attracted new residents. This is also where it is most difficult to increase the housing stock since central areas of this region are already heavily built-up. However, even in close proximity to the city centre there are opportunities for expansion, e.g. in former industrial areas, while the stock of flats could in principle also be increased by constructing more multi-storey residential buildings.

In Denmark residential construction is subject to the restrictions of land zone legislation regulating where new dwellings may be built. There are still large areas of Greater Copenhagen with rural zone status, cf. Chart 5, that to some extent could be included in the urban zone and used for new construction, subject to appropriate infrastructure expansion as necessary.

Housing consumption
The average dwelling size has grown over the last 25 years. On a national basis for all types of dwelling the average dwelling size today is 110 m². At the same time the number of occupants per dwelling has fallen, thereby increasing the number of m² per person from 48 to 57, equivalent to an increase of almost 20 per cent, cf. Table 1.

Distribution by municipality groups shows a clear connection between housing consumption and the price of that consumption. The lower the cost of one square metre of living space, the higher the number of m² demanded, cf. Box 3. Today a resident of a rural municipality (municipality group 8) has an additional 15 m² of living space at his disposal compared to a resident of the City of Copenhagen or the City of Frederiksberg (municipality group 1), and the gap has widened in the last 25 years, cf. Chart 6. The number of occupants per dwelling is lower in group 1 than in the rest of the country, but this is more than offset by the fact that the dwellings are typically also smaller, so that the number of m² per person is the lowest among all municipality groups.

Housing prices and housing wealth
Housing wealth in Denmark can be calculated on the basis of housing stock statistics supplemented with data from the Association of Danish Mortgage Banks' real-property price statistics covering traded prices per m² for one-family dwellings, freehold flats and holiday dwellings. Housing wealth can be calculated in different ways, depending on the delineation of ownership and use of the housing stock. Total housing wealth gives an estimate of the market value of all dwellings. Naturally this calculation is subject to uncertainty since traded prices for e.g. rented dwellings, including social housing, are not immediately available. However, market prices for owner-occupied dwellings can be taken as a rough estimate of the relevant prices. In addition, an estimate of the households' housing wealth can be calculated. This is the market value of the dwellings that are owned by private individuals^[2], which is e.g. relevant to an assessment of the households' total wealth. In addition, it is possible to calculate owner-occupied housing wealth, comprising the household-owned dwellings that are owner-occupied. This omits the dwellings rented out by households.

At the beginning of 2005 housing wealth totalled kr. 3,500 billion, of which almost half was the value of dwellings occupied by the owner, cf. Appendix Table 2^[3]. On measuring the regional distribution of housing wealth it is relevant that the cost of a home is significantly higher in large towns and cities than in smaller urban areas, cf. Chart 8. The Greater Copenhagen Region and the large provincial towns (municipality groups 1, 2 and 4) thus account for half of the total value of all dwellings for permanent residence, but only one third of the equivalent living space.

The net effect on the price of housing of a smaller living space, but a higher price per m² in the large towns and cities compared to small towns is shown in Table 2, in which the average value per dwelling in the various municipality groups is calculated. There are large regional variations in the average values of the homes. A detached one-family house situated in the Greater Copenhagen Region is on average almost three times as expensive as in an average rural municipality. So there are large variations in opportunities to get onto the housing ladder with a given income. However, average incomes also vary in size across municipality groups and the higher prices in the large cities are partly due to the fact that this is also where the highest incomes are to be found.

In addition to variations in the number of m² there can also be regional variations in the quality of housing. Despite major urban renewal during the past decade the City of Copenhagen is still top of the league in the statistics for dwellings lacking essential installations such as central heating and own WC and/or shower, just as a relatively large proportion of the homes are disturbed by noise. So there is a lot of variation in the quality of the homes.

The number of households has increased during the last 10 years, and especially the number of single occupants has risen, cf. Table 3. Distribution by type of dwelling and family type shows that single occupancy of multi-family buildings dominates. This has contributed to increased demand for flats and thereby to especially high price increases for freehold flats. Many single people live in the Greater Copenhagen Region, which has contributed to strong price pressure in this region.

Taxation of owner-occupied dwellings
As stated, the high prices per m² in the Greater Copenhagen Region entail that a relatively large share of Denmark's total housing wealth is located in this region. This also applies to owner-occupied housing wealth, which solely comprises dwellings occupied by the owner. The last-mentioned is the relevant definition of wealth in relation to property value taxation since dwellings that are rented out are not subject to property value tax.

The uneven regional distribution of owner-occupied housing wealth is also apparent from housing tax paid. Distribution of property value taxation revenue – previously the revenue from taxation of rental value – on municipality groups shows that a large and growing share of the revenue stems from the Greater Copenhagen Region broadly defined (municipality groups 1 to 3), cf. Chart 9. Box 4 presents an overview of property taxation in Denmark.

Relating the property value tax revenue to owner-occupied housing wealth in the individual municipality groups gives an effective property value tax rate at municipality group level, cf. Charts 10 and 11 and Appendix Table 4. Holiday dwellings are omitted from the wealth statistics at municipality group level in order to make adjustment for the fact that the revenue from the taxation of holiday dwellings falls to the owner's municipality of residence. Box 5 presents a description of the method used to calculate the effective property value tax rate.

The effective property value tax rate has been falling since the beginning of the 1990s. The decline was greater during the 1990s than it has been under the tax freeze from 2002 up to the present day. However, in its present form the tax freeze will cause the effective property value tax rate to approach zero in the long term, since the revenue does not keep up with increases in housing prices. Even though an increasing share of the property value tax revenue stems from the Greater Copenhagen Region it is nonetheless in this region that the effective property value tax rate has declined most.

Revenue data is only available up to the end of 2004. For 2005 and 2006 the property value tax revenue is estimated by assuming an increase in the annual revenue equivalent to the increase from 2003 to 2004. Despite the tax freeze, the revenue still increases, which is a consequence of an influx of new dwellings, including extensions, and because the special reduction of 0.2 per cent lapses when a dwelling is sold.

Based on the above, the effective property value tax rates for 2005 and 2006 can be estimated at respectively 6 and 5 per thousand, i.e. 0.6 and 0.5 per cent. The change in the effective property value tax rate is dominated by changes in housing prices, so that the calculated effective tax rates for 2005 and 2006 must be expected to be reasonably precise.

Even with a formal property value tax rate of 1 per cent the effective tax rate will usually be lower since typically a dwelling will be valued below the market price, just as there are various types of reduction, cf. Box 4. The spread between the property valuation and the market price will typically widen in a soaring market, which is one reason for the drop in the effective tax rate in the second half of the 1990s. In addition, there is the reduction of the rental value rate from 2.5 to 2.0 per cent as from 1994.  

Taxation of real property also includes the land tax (" grundskyld" ). There are no immediately available statistics for revenue by the property category used, i.e. dwellings occupied by the owner, since land tax is also paid on rented dwellings and on agricultural land.

The revenue from land tax on owner-occupied dwellings can be estimated to be equivalent to the revenue from property value tax, i.e. around kr. 10 billion in 2005. Land tax revenue increases as land prices rise since the tax freeze does not include the taxation basis (land values) in the same way as for property value taxation. The taxable land values rose by more than 20 per cent from 2002 to 2005, which was less than the increase in total property values. In 2006 the tax authorities in many cases undertook substantial upward revision of taxable land values. Since the tax freeze limits the increase in property taxation from year to year a large increase in land values will push up taxation not only next year, but for a number of years, until the tax increase has taken full effect.

APPENDIX

Municipality groups of the Danish Customs and Tax Administration (Skat)

Municipality group 1:
Copenhagen and Frederiksberg

Municipality group 2:
All municipalities in the County of Copenhagen  

Municipality group 3:
All municipalities in the Counties of Frederiksborg and Roskilde

Municipality group 4:
Other municipalities with 50,000 residents or more in the largest built-up urban areas:
Århus, Odense, Ålborg, Esbjerg, Kolding and Randers

Municipality group 5:
Other municipalities with 20-50,000 residents in the largest built-up urban areas

Municipality group 6:
Other municipalities with 10-20,000 residents in the largest built-up urban areas

Municipality group 7:
Other municipalities with 5-10,000 residents in the largest built-up urban areas

Municipality group 8:
Other municipalities with less than 5,000 residents in the largest built-up urban areas