interpretation of the results difficult. They are, however, shown in Table 5, but without any
comments in the text.
As far as total migration is concerned, the explanatory value has increased throughout the
entire period and the variables included in the regressions encompass the causes for “job
migration” relatively well. The net migration, with regard to studying, is also encompassed
relatively well by the “explanatory” variables. The labour market and work force related
variables appear to also influence net migration to and from studying; study migration can
naturally be regarded as a work-force related migration since it is an investment in human
capital. Furthermore, the labour market in the college and university towns, for example, is
characterised by the very fact that they are study locations.
As far as the input variables are concerned, we see that population size (POP) as a rule has a
positive effect on net migration. The larger the local labour market is, the greater is the net in-
migration, relatively speaking. This also means that small local labour markets as a rule have
a net out-migration, at least when viewed overall. For the various categories we also find the
same pattern concerning migration to and from work in 1991 and 1997. Both of these years
can be seen as relatively good years from a labour market point of view. This should,
however, not be interpreted as implying that good years mean that external recruitment
increases with an economic upturn. Instead the opposite appears to be the case. On the other
hand, the excess in-migration to regions with a large population and good employment
situation increases.
The employment situation (EMP) also has the greatest impact on those who migrate to and
from work. The better the employment situation, the greater the excess in-migration of work-
related migrants and vice versa. This applies to all the years, i.e. the economic situation does
not appear to alter this pattern. As far as the unemployed are concerned, we find the same
correlation in 1996 and 1997, while this does not hold for the years 1991 and 1993, when the
economy took a downturn. It should, however, be kept in mind that we do not know the
destination, and that a person may well have had a job when he or she moves in or out but
then lost it during the year when the economic situation worsened. The local turnover
(LOCTURN) of the work force has, however, not had any influence on the net migration,
either overall or with regard to “job migration”.
14