spatial category. The commuting flows into agglomeration centers and into urban fringes are
very low.
The illustrations (Fig. 3, Fig. 4) should point out that there is a connection between employee
density and commuting behaviour. To answer this question the employee concentration is put
in relation with the commuting rate and the average commuting distance of the 440 German
districts (NUTS 3) distinguished in relation to spatial category. Fig. 3 shows a clear
connection between employee density and commuting rate. With decreasing employee
density commuting rate rises. The differences between the spatial categories are clearly
visible. The agglomeration centers show with a high employee density the lowest commuting
rates. They concentrate in the left upper area of the illustration. In contrast the employment
density in the urban fringes is lower although it is still at a high level. The commuting rates
vary between 30% and 60%. The majority of low density and peripheral regions concentrate
in the bottom part of the illustration. Even though their employment density is low, their
commuting rates are lower than in the agglomeration centers.
In contrast the connection between employment density and average commuting distance is
weak. With decreasing employee density the average commuting distance rises slightly (Fig.
4). Indeed, the above described connections between spatial categories can be discovered,
however, in total spatial categories are more strongly scattered.
The explanations have shown that there are spatial differences in commuting behaviour.
Commuting rates as well as average commuting distances are varying spatially. Frequently
they fall together with special spatial categories. Also the consideration of commuting flows
shows spatial differences in commuting behaviour. It was shown that the supply of jobs, as a
possible spatial explanation variable, has a strong effect on the explanation of the commuting
rate. The higher the employee density in a region the lower the commuting rate. This applies
in particular to agglomeration centers and low density regions. However this connection is
weak for urban fringes and peripheral regions. Thus even urban fringes with high employee
density show quite high commuting rates. The reason for this is the immediate spatial
neighbouring to agglomeration centers. It was shown that the majority of employees from
urban fringes are oriented towards agglomeration centers.
The element of employee density as a means of explaining the average commuting distance is
weak. The average commuting distance increases with decreasing employee density in most
spatial categories. For agglomeration centers and peripheral regions this connection is not
valid. In these regions the average commuting distance increases with employee density.
Different commuting distances seem not to be determined by spatial structure. Another
explanation is that employees that did not find a suitable job in the agglomeration centers are
forced to search beyond the boundaries of their region. This can be put down to the fact that
most employment opportunities still can be found in the agglomeration centers. In comparison
low density and peripheral regions are offering much lower employment opportunities.
Because of that employees have to extend their search to other agglomeration centers. Rather
it seems that there must be more reasons to be taken into consideration, for example special
regional circumstances like polycentric or monocentric spatial characteristics. Moreover
reunification processes could explain differences in commuting distance between West and
East Germany to a certain degree. But in this paper they should not be an object of
investigation.