What is the BR-index and what does it perform?

BR (building and renovating) is a statistic measure of the market potentials in the building and renovation sector. Data basis are the statistic data on area level and on 5 digits Postal code level. BR shows the positive and negative deviations from the market potential of the respective country (= 100 %).

 Thus the BR- index is an instrument for:

·         computation of regional market potentials

·         spatial optimization of the distribution structure

·         logistic planning

Although a complicated mathematical model stands behind BR, the computation of the market potentials with the available key figures is very simple.

How does the BR-Model work?

The model, which BR is the basis, is a so-called "nonlinear multiple regression approach" (see Toolbox 1). This procedure looks for a relationship between the market potential and its statistic factors of influence. This relationship is held in form of an equation, which became compiled at a well-known sample (calibrated).

The market potential of course depends on a very large number on statistic influencing factors. But not all factors cannot taken into consideration by computating the BR. It requires thus a data reduction, in order to receive the most important influencing factors. It is particularly important that several factors of influence, which exert the same influence are replaced by a factor. The statistic procedure of this data reduction is called "factor analysis" (see Toolbox 2).

Thus there are influencing factors, at which others are seen statistically as "attached", without being considered in detail. That means the actual computation factors of BR still consider another multiplicity of other influences, which are not explicitly specified.

The analysis factors, which are the basis for the BR, are: ·

·         Building permits: Residential building

·         Building permits: Dwellings

·         Building permits: Construction costs

·        Dwelling size

·         Land development structures

·         Housing stock

·         Purchasing power (GfK, Nürnberg)

·         Population density

·         Employees industry

·         Employees furniture industry

        ·         other.

The BR model does not work additive, e.g. as a simple multiple linear regression model. Thus BR considers the purchasing power factors and turnovers in the form of delimitation factors of the construction costs data, related to the number of building permits, which again are standardized on the dwelling structure.

What BR cannot perform

Exceptions do not confirm a rule, but they draw our attention to the borders of our model. The quality of a rule depends primarily on the number of exceptions. If this is small, i.e. the error probability small, then the rule is good. Like all statistic models so BR also has an error probability (see Toolbox 3). However, this is small with a=0.09 (in the f-test) for a socio-demographic model. With such complex systems such as WBR it describes, there is no 100 %-ige explanation given.

A further delimitation of BR lies in the kind of the input data, which among other things are related to building permits and the relative construction costs. That means, building removals and renovations, which do not require a building permit, are not seized directly. Indirectly however a coupling exists, because a high building activity subject to permission usually correlates with building activities not subject to permission.