Skip to content

Accuracy and reliability

Contact info

Prices and Consumption, Economic Statistics
Jakob Holmgaard
+45 39 71 31 24

jho@dst.dk

Get as PDF

Sales of real property

There is no sampling error in the final figures, but the price indices are still subject to some degree of uncertainty. This is due to the fact that the same houses are not sold in each period which makes it necessary to control for quality changes between sold houses. No matter how this quality correction is done it will always contains some degree of uncertainty. The statistical uncertainty is explained here. Bias on the preliminary price indices is close to 0, while the average deviation is about 0.3 for the quarterly figures.

Overall accuracy

The final figures contains almost all transactions of real estates and the number of sales is therefore likely to have a high precision. However, the final price indices are subject to some degree of uncertainty. This is due to the fact that there is variation in what properties there are transacted in each period. So even though the actual sales prices are known, the price development depends on the quality and number of transacted properties. This leads to a statistical uncertainty, which is important to take into consideration when using this statistics. The quality correction is done by calculation the spar-value for each single property, which is the price divided by the appraisal (Sales Price Appraisal Ratio-method). However, the spar-value has two limitations: Firstly, two identical properties with the same quality and comparable location is not necessary sold at the same price at a certain time, since the market is not perfectly transparent. For example can one seller get a higher price if he can wait for the 'right buyer'. There is always an element of uncertainty in the transacted price. Secondly, two identical properties does not necessarily have the same appraisal-value. The preliminary figures are subject to some uncertainty. This is due to the fact that the calculation is done before all transactions have been registered into the electronic land registration system. The actual period is based on approximately 90 percent of all transactions in this period. The second time the period is published approximately 95 percent of all transactions is included in the calculation. Bias shows if the revisions of the price indices in averages is under- or overestimated. Bias is calculated as the average of the monthly deviations in percentage points. Deviation shows how close the first publication of a period is to the final version. Bias and deviation of price developments are described throughout the published New articles.

Sampling error

This statistic is a total statement of all registered real estate transactions based on administrative register information.

Non-sampling error

The final figures are based on almost all transactions. However, price indices are subject to statistical uncertainty. This is due to the fact that there is variation in what properties that are transacted in each period. So even though the actual sales prices are known, the price development depends on the quality and number of transacted properties. The uncertainty of the quarterly price indices at the nation wide level is +/- 1.1 percentage points for one-family houses and +/- 1.7 percentage points for owner-occupied flats. The uncertainty is higher for owner-occupied flats than for one-family homes because the number of transactions are lower. The uncertainty is higher at regional level than at the nationwide level because the number of transactions are lower. The statistical uncertainty for house price indices is explained here .

The preliminary figures are revised for two reasons: Firstly, there can be sampling errors since the calculations is done before all transactions have been registered into the electronic land registration system. Secondly, the base year of the appraisals is changed once a year in order to include the newest information of changes in the property stock, for example new buildings.

Quality management

Statistics Denmark follows the recommendations on organisation and management of quality given in the Code of Practice for European Statistics (CoP) and the implementation guidelines given in the Quality Assurance Framework of the European Statistical System (QAF). A Working Group on Quality and a central quality assurance function have been established to continuously carry through control of products and processes.

Quality assurance

Statistics Denmark follows the principles in the Code of Practice for European Statistics (CoP) and uses the Quality Assurance Framework of the European Statistical System (QAF) for the implementation of the principles. This involves continuous decentralized and central control of products and processes based on documentation following international standards. The central quality assurance function reports to the Working Group on Quality. Reports include suggestions for improvement that are assessed, decided and subsequently implemented.

Quality assessment

Properties that are sold at different times are different in terms of size, location and standard. In order to calculate the pure price development between different periods Statistics Denmark uses a method which adjusts for some of these quality differences. For each property the sales price is divided by the appraisal, which is called the 'spar-value'. It is this spar-value that is used in the calculation of the price indices and not just the pure sales price. The method is internationally accepted and is known as the SPAR-method (Sales Price Appraisal Ratio). The SPAR-method is also used in Sweden and the Netherlands.

Analysis made by Statistics Denmark shows that the transactions of properties is typically registered into the electronic land registration system close to the acquisition date, which is the time when the new owners gets the key to their new property. The acquisition date may be relatively long time after the transaction date. The analysis also shows that the period between the transaction-date and the acquisition date is longer for more expensive properties than for cheaper properties. At the same time analysis also shows that expensive properties is more likely to have a relatively lower appraisal than cheaper properties. That is, the appraisals for expensive properties is typically less than the sales price while it is opposite for the cheaper properties where the appraisals typically is higher than the sales price. The preliminary price indices are calculated in a way that adjust for these circumstances in order to keep the bias and deviation as low as possible. The method is described in detail: here.

Data revision - policy

Statistics Denmark revises published figures in accordance with the Revision Policy for Statistics Denmark. The common procedures and principles of the Revision Policy are for some statistics supplemented by a specific revision practice.

Data revision practice

All figures are revised 11 times. The calculation is made on the basis of the transactions which have been registered in the electronic land registration system at the time of calculation. The preliminary figures are calculated based on an incomplete set of data since not all transactions have been registered at the time of calculation.