Statistical processing

Source data

Most data used in the construction cost index for residential buildings is gathered from other statistics at Statistics Denmark. Prices of materials and equipment are collected from the Price Index for Domestic Supply. Around 950 different material prices are collected every month from producers and importers, cf the documentation of statistics for Producer and Import Price Index for Commodities. Prices are also gathered from Molio Prisdata.

Prices of transportation are collected quarterly from the Producer price index for services, from which prices of Freight transport by road and removal services are used.

The prices of total labor costs are mainly collected from the implicit indices of average earnings. The prices are collected each quarter and refer to the second month of the quarter, cf the documentation of statistics for Implicit index of average earnings.

Weighing of different cost units in the construction cost index is based on three concrete buildings: One one-family house, one semi-detached house, and one multi-family house. The buildings were chosen as representation of typical housing constructions on the basis of an analysis of actual housing constructions during the weight reference period (2015), primarily on the basis of the Dwelling- and Housing Register (BBR) . The data for the three types of buildings are obtained from contractors of different size and geographic location. A consulting contractor specialized in pricing of housing construction was involved in the processing of the data.

Frequency of data collection

Primarily monthly and quarterly. Some prices are only gathered on a yearly basis.

Most prices used in the calculation of the index are three-month averages for the price of materials or services on the 15th day of each month in the quarter. If the price of a material or service on the 15th day cannot be gathered, the price is gathered from a suitable day earlier in the month, or an average of prices in the first 15 days of the month is used. The three-month average price is considered representative for the entire quarter.

Data collection

Material prices are collected for the Price Index for Producer and Import Price Index for Commodities (PPI) via Virk, and are reported by select, representative companies. The Construction Cost Index for residential buildings uses PPI-prices belonging to NACE Rev. 2 Section F. There are about 950 material prices forming about 120 representative material goods. Furthermore, some material prices are gathered manually from Molio Prisdata. Prices representing freight transport are gathered from the Producer price Index for Services, and are also collected via Virk.

Labor Costs are collected for the implicit indices for average earnings. These data are collected from IT-systems used by reporting companies to administer employee salaries on an individual level. The data is mostly reported system-to-system, e.g. bulk data from dedicated salary IT-systems reporting on behalf of their clients to Statistics Denmark. Companies that have their own salary IT-system report data either by upload via a web application or by sending encrypted files via email to a dedicated email address. The Construction Cost Index uses earnings for NACE Rev. 2 Section F.

Data validation

The data validation is handled by the Price Index for Domestic Supply and the implicit indices of average earnings. The first validation of price data happens when prices arrive. Here they are tested for unusual changes. Prices that do not pass fixed threshold values will be checked manually by the staff and accepted only if the companies can verify the change.

When all prices are received, the system generates a list that includes all price changes and a measure of how these affects the elementary aggregates. The last validation is a visual inspection of all index tables.

Data compilation

The Construction cost index for residential buildings consists of a hierarchical system, where the collected prices are first divided into a number of product groups. The groups are aggregated to base aggregates from which base prices are calculated as geometric averages. Base prices are used to calculate base indices, which show the development of base prices in successive time periods. Base indices constitute the most detailed index level. On the basis of base indices, aggregated subindices are done using Laspeyres type index calculations. Finally, aggregated total indices are calculated, which are the top level of the index hierarchy.

Pricing weights in the Construction cost index for residential buildings are based on classification codes and weights for the different types of housing constructions. While product prices are updated quarterly, weights are updated approximately every five years.

Weights of different cost units in the construction cost index are based on accounting data from three actual building constructions: One one-family house, one semi-detached house, one multi-family house. The buildings were chosen as representation of typical housing constructions on the basis of an analysis of actual housing constructions during the weight reference period (2015), based primarily on data from the Dwelling- and Housing Register (BBR). The data for the three types of buildings are obtained from contractors of different size and geographic location. A consulting contractor specialized in pricing of housing construction was involved in the processing of the data.

The accounting data used is comprised of different production parts, e.g., assembly of a concrete wall, which is divided into costs of involved materials (e.g., concrete), equipment (e.g., crane), and profession (e.g., concrete slab work). This is the basis of price weights. Costs of usage and operation of construction sites, including winter specific measures, are distributed proportionally to the different professions and materials.

Adjustment

Substitution of the products used in calculation of the index may occur over time. Because such new products may be of a different quality than the previous, continuous quality control of is performed.

No other data corrections are carried out, except for the methods described in the data validation section.