This page is archived.

Data published after 5 April 2022 can be found on the renewed website.

Go to the new statistics page

Quality Description

1. Relevance

1.1. Information content and purpose of us e

Quarterly statistics on housing prices describe the unencumbered prices per square metre of old and new dwellings in housing companies, and year-on-year and quarterly changes in them. The statistics contain data classified by area - old dwellings also by sub-area - type of building and number of rooms from the examined quarter and from a longer period of time. The purpose of the statistics is to provide information about developments on the housing market to all interested parties.

1.2. Concepts, classifications and data

The data and the data suppliers

Old dwellings : The data of the statistics on dwelling prices are based on the price information gathered by the National Board of Taxes for asset transfer tax calculation purposes. The real estate register of the National Board of Taxes and Statistics Finland’s data file on the dwelling stock, which is derived from the Population Register Centre’s register of buildings and dwellings, are also exploited as data sources for these statistics.

New dwellings : The data of the statistics on dwelling prices are based on information Statistics Finland re-ceives via a private price monitoring service about transactions in new dwellings made by the largest real estate agents and building contractors

Used concepts:

Dwelling: A dwelling refers to a room or suite of rooms that is equipped with a kitchen, kitchenette or cooking area and is intended for year-round habitation.

Price per square metre of dwelling : The statistics are compiled from data on unencumbered prices, in other words prices inclusive of debt portion. The published price concept is price per square metre (€/m2).

Floor area of dwelling : The floor area (m2) of a dwell-ing is calculated from the inner surfaces of the walls enclosing it. The floor areas of auxiliary spaces (utility space, walk-in wardrobe, etc.), bathroom, hobby room, fireplace room, sauna in dwelling, washroom and changing room, and rooms used as working space if no hired employees work in them are also included in the floor area of a dwelling. Garage, cellar, sauna space in unoccupied basement, unheated storage space, balcony, porch, veranda, vestibule and unoccupied attic space are not included in the floor area of a dwelling.

First home : First home transactions refer to the transac-tions entitled to first-time homebuyer’s exemption from the asset transfer tax (www.vero.fi)

Old/new dwelling: An old dwelling refers to a dwelling that has not been completed in the examined year or the year before it. Respectively, a new dwelling refers to a dwelling completed in the statistical reference year or the year before it that is sold for the first time.

Type of building : The dwellings in the statistics are classified into blocks of flats and terraced houses. The data on terraced houses also cover detached houses whose tenure is based on ownership of housing com-pany shares.

Type of financing: Dwellings financed with ARAVA subsidised housing loans and price controlled HITAS dwellings are not included in the non-subsidised dwelling category used in the statistics.

Number of rooms: A room is defined as a space with one or more windows that has a floor area of at least seven square metres and mean height of at least two metres. A hall, porch, bed recess or other similar space is not regarded as a room. Kitchen is not included in the number of rooms. Dwellings with at least three rooms are classified into room number category 3+.

(Nominal) price index: Describes price change compared to the base year (old dwellings 2005, 2000,1983 or 1970 and new dwellings 2005) of the index concerned.

Real price index : Describes real price change compared to the base year (old dwellings 2005, 2000,1983 or 1970 and new dwellings 2005) of the index concerned. Real price index is calculated by dividing the point figure of the nominal price index by the point figure of the consumer price index of the corresponding point in time and base year.

Distribution parameters :

Q1 (lower quartile) = 25% of the prices per square metre are lower than or equal to the lower quartile.

Med (median) = Middle price of prices per square metre arranged in size order.

Q3 (upper quartile) = 75% of the prices per square metre are lower than or equal to the upper quartile.

Classifications:

Regional division, old dwellings : The statistics use diverse area combinations, such as Greater Helsinki Area, satellite municipalities around the Greater Hel-sinki Area, regions and urban sub-areas. The Greater Helsinki Area comprises Helsinki, Espoo, Vantaa and Kauniainen, which in statistics is included in Espoo. The satellite municipalities are Hyvinkää, Järvenpää, Kerava, Kirkkonummi, Nurmijärvi, Riihimäki, Sipoo, Tuusula and Vihti. Regions are defined according to the decision of the Council of State of 26 February 1998. The urban sub-areas are formed of postal code areas using price level and location as the criteria. Details of the used regional classifications are appended to this publication and can be found on Statistics Finland’s website.

Regional division, new dwellings: Due to the low number of transactions statistics on the prices of new dellings are compiled according to less detailed regional division than statistics on the prices of old dwellings. The classification used in the statistics on the prices of new dwellings also takes into consideration the needs of the Consumer Price Index, hence the regional classifi-cation uses the division into major regions. The area categories are (1) Whole country, (2) Greater Helsinki Region (same as with old dwellings), (3) Rest of Fin-land (Whole country exclusive of Greater Helsinki Region), (4) Rest of Uusimaa (exclusive of Greater Helsinki Region), and Itä-Uusimaa and major regions: (5) Southern Finland, (6) Western Finland, (7) Eastern Finland and (8) Northern Finland.

2. Methodological description

The calculation method of the indices for old dwelling prices 2000=100 and new dwelling prices 2005=100 combines the classical classification approach and regression analysis (so-called hedonic method). The index aims at answering the question how much more/less a typical dwelling in a housing company costs now compared to before. Monitoring average price changes will not necessarily provide an adequate answer, since average prices change also because the composition of dwellings sold at different times varies. For example, the relative shares of different types of dwellings vary from quarter to quarter.The method aims at distinguishing better than be-fore the true price developments from price effects arising from compositional changes.

Because location, type of building and number of rooms are the most important price determinants, the composition of sold dwellings is first standardised for these variables by classification. The regional classification has been constructed so as to be geographically meaningful and as homogeneous as possible in respect of price levels. In the statistics on old dwelling prices the largest municipalities are divided into several sub-areas, and the smallest municipalities where few transactions take place have been combined. In the statistics on new dwelling prices the regional classification has been formed according to six sensible geographical entities because due to the low number of observations in the data a more detailed classification cannot be used. In respect of both old and new dwellings, the dwellings within an area have been stratified by type of building into dwellings in blocks of flats, and dwellings in terraced and detached houses. Dwellings in blocks of flats have been classified further by number of rooms into dwellings with one room, dwellings with two rooms and dwellings with three or more rooms. Old dwellings in terraced houses have been divided by number of rooms into two categories — dwellings with fewer than, and dwellings with at least three rooms. New dwellings in terraced houses form one category.

The used classification does not necessarily homogenise the data sufficiently. Factors affecting price, such as micro-location, floor area, year of completion, and so on, are not controlled for by the classification. The available data contains information on these characteristics, which can be used for adjusting the average price of a given category in the comparison period so that the obtained average price adjusted for quality takes into account compositional changes within the category in the base and comparison periods. The following regression equation model are specified:

Regression model for average square metre

Regression model for average square metre

The notation of model is standard is the logarithmic price per square metre of dwelling floor area of dwelling i in location j Variables is the micro area indicator (postal colde areas in large urban centres and municipality indicators in combination areas). In model 1 the variable ’huone’ indicates the number of rooms, RT indicates terraced house dwelling and (RT)*(huone3) is the interaction term for a terraced house dwelling with at leat three rooms..

The models are estimated using ordinary least squares (OLS) for each location separately. The models were not estimated for each class, because this would have lead to degrees-of -freedom problems. The functional form is standard semi-log and the square roots of dwelling floor area and, in respect of old dwellings, of construction year are included as explanatory variables to capture non-linear price effects. The ‘huone’ indicators are naturally strongly correlated with dwelling-floor area, but they are included for technical reasons, namely in this way it is guaranteed that the sum of residuals in the base period (year 2000) are zero in all index classes.

Let denote estimates of the model parameters,  in the index class of i by vector

Estimaton vector of model in class i

Estimaton vector of model in class i
and the sample characteristics (construction year, size, postal code area indicator) of the dwellings in the base and comparison periods respectively.

Average price vector for base period

Average price vector for base period

Average price vector for comparison period

Average price vector for comparison period
Then within each class the quality adjustment due to differences in construction year, dwelling floor area and location according to postal code area can be written as:

Quality adjustment in index class i

Quality adjustment in index class i
The quality adjustment works in the following way: If, for example, the average construction year of old dwellings is older in the comparison period than in the base period, the index must be corrected upwards, because otherwise lower prices due to earlier construction year would be wrongly interpreted as price fall. The size of the adjustment depends on the difference in the average construction year of the dwellings and on the estimated construction year coefficients in the regression.

The overall index point-number for the whole country is obtained via aggregated price changes in every index class and price adjustmentst so callede log-Laspeyres formula

Log-Laspeyres Index Formula

Log-Laspeyres Index Formula
In model (2) is N number of index classes,

Geometric price ratioiin class i

Geometric price ratioiin class i
and

The weight in clas i

The weight in clas i
Geometric prices are calculated for observations' prices per square metre via the following formula:

Geometric average prices

Geometric average prices
The weights for old dwellings are derived as value-shares of the stock of apartments in 2005.

Calculation of weights

Calculation of weights
,where

Average dwelling-floor area of the dwellings in class i in year 2005

Average dwelling-floor area of the dwellings in class i in year 2005
and

Number of dwellings in the class concerned

Number of dwellings in the class concerned
and

Average price of class concerned in year 2005

Average price of class concerned in year 2005

3. Correctness and accuracy of the data

3.1. Reliability of the statistics

The statistics on the prices of old dwellings are based on the asset transfer tax data of the National Board of Taxes, which cover the transactions of all dwellings whose tenure is based on ownership of housing company shares. All transactions of old housing company dwellings are not included in the statistics, because the purchaser is allowed two months to pay the asset transfer tax. Many purchasers pay the tax more quickly than this and in transactions intermediated by real estate agents the tax is paid at the time of transaction.

When the statistics are published they cover approximately two-thirds of all transactions made in the latest statistical reference quarter. Statistics Finland receives the data on the remainder as they arrive at the National Board of Taxes. The quarterly data are updated retrospectively so that the final data for a given year are published with the data for the first quarter of the year following it.

The statistics describe the housing company share market by area relatively reliably. However, the number of included transactions should be taken into consideration. If few transactions have been made, a couple of deviating cases may affect the average price for an area significantly.

The statistics on the prices of new dwellings are based on data obatained from the largest real estate agents and building contractors and is a final when first published.

3.2. Accuracy of the statistics

Cases with missing information about transaction price or floor area, or with exceptionally high or low price due to contract within family or error in data entry are not accepted into the statistics. The acceptable ranges of prices per square metre in statistics 2008 and 2009 are: €/m2 1,200–9,000 for the Greater Helsinki Area, €/m2 800–6,500 for Tampere, Turku, Jyväskylä, Kuopio, Oulu, Vaasa and the satellite municipalities surrounding the Greater Helsinki Area, and €/m2 500–5,000 for other areas.

Confidence interval of 95% has been calculated with the bootstrap method for the housing price index of old dwellings. For the whole country, the confidence interval is ± 0.7%, for the Greater Helsinki Area ± 1.4% and for the rest of the country ± 0.8%.

3.3. Use of the parameters of the statistics

Because the index takes into account changes in the distribution of year of completion (for old dwellings only), floor area and location of dwellings sold at different points in time, and their effects on prices, the average prices of the statistics vary differently from the price index. This has been done because the price index and the average price are each useful measures for different situations.

The price index endeavours to measure as accurately as possible how much more/less an average dwelling in a housing company costs now than it did before. The aver- age price , in turn, describes the prevailing price level for sold dwellings without considering whether they are older, newer, larger or smaller than dwellings sold be-fore.

4. Timeliness and promptness of published data

4.1. Publication frequency and measurement period of the statistics

Quarterly statistics on housing prices are compiled per quarter and published one month from the end of the examined quarter

4.2. Preliminariness of the statistics

When the statistics are published they cover approximately two-thirds of all transactionss in the latest statistical reference quarter. Statistics Finland receives the data on the remainder as they arrive at the National Board of Taxes.

The quarterly data are updated retrospectively so that the final data for year t are published with the data for the first quarter of the year following it.

5. Accessibility and transparency of the data

A latest data release will be published from the statistics on Statistics Finland’s website on the publication date of the quarterly statistics on dwelling prices. The entire publication can be ordered as a printed paper version or an electronic pdf version. Data concerning dwelling prices can also be found from Statistics Finland’s web pages and database service.

The essential metadata have been described in this document, which is incorporated into the quarterly publi-cation of statistics on dwelling prices. This document is also available on Statistics Finland’ web pages.

This statistics covers only dwelling transactions in housing company shares. Especially in the Greater Helsinki Area, there are numerous real estate transac-tions that are not included in these statistics. Data on real estate transaction prices by municipality are avail-able from the National Board of Survey (Tel.: +358 40 801 1204).

6. Comparability of the statistics

6.1. Comparability with other data

When these statistics are compared with data from other producers the source of the basic data should be considered. Statistics Finland’s data derive from com-prehensive files of the National Board of Taxes, and thus cover exhaustively all completed transactions.

6.2. Comparability over time

Statistics compiled from the asset transfer tax data of the National Board of Taxes and classified according to these current quarterly statistics are available on the prices of old dwellings starting from the year 1987.Older data are available going back to 1970. The statistics for the 1970 to 1986 period are based on data provided by real estate agents and the used classification is much less detailed than the one used since 1987. For the prices on new dwellings time series have been calculated since 2002.

7. Coherence and consistency

Statistics Finland published prices statiscs of corporation flats.and price statisscs of real estate prices quarterly. Besides the data published by Statistics Finland, real estate agents, credit institutions and banks also publish information concerning dwelling prices and their development. More on differences between the published data under section 6.1. above..


Source: Prices of dwellings, Statistics Finland

Inquiries: Petri Kettunen (09) 1734 3558, Paula Paavilainen (09) 1734 3397

Director in charge: Kari Molnar


Updated 30.07.2009

Referencing instructions:

Official Statistics of Finland (OSF): Prices of dwellings in housing companies [e-publication].
ISSN=2323-8801. 2nd quarter 2009, Quality Description . Helsinki: Statistics Finland [referred: 28.3.2024].
Access method: http://www.stat.fi/til/ashi/2009/02/ashi_2009_02_2009-07-30_laa_001_en.html