European Regional Science Association
|
European Regional Science Association
|
|
|
|
|
The abstract for paper number 468:
Gerke Hoogstra, Jouke Van Dijk, University of Groningen, Groningen, The Netherlands
Specification issues and sample selection in population-employment interaction models
In this paper, a meta-regression of analyses on selected Dutch localities (1988- 2002) is presented to shed light on the ambiguity that surrounds the issue of causality between population and employment growth across space (i.e. do jobs follow people or do people follow jobs?). The use of primary analyses enables experimenting with different applications of a population-employment interaction model in a controlled set-up. From these analyses 2,772 estimates are obtained that reveal the jobs-people direction of causality for different model specifications and different samples of the data. The meta-regression model tests for the impact of employing levels versus densities as the variable measurements, spatial weights matrix specification and the effect of time period, region type and employment type under investigation.
Unfortunately full paper has not been submitted.