Introduction

The addition of the spatial dimension in regional modelling enlarges the need for data, whereas regional data are usually much less readily available than national data. The econometric time-series approach, which is the common approach in all macro-economic models, cannot be used when regional time series are not available. Hence, it is not surprising that many regional model builders have concentrated their efforts on the cross-section alternative, which aims at getting a detailed description of the regional economic structure at a single moment in time. As a consequence, sectoral disaggregation and input–output analysis have been more popular among regional economists than among their macro-economist colleagues (Rose and Miernyk, 1989).

A second distinctive feature of regional models is the prominence that labour market modelling takes. This is directly related to the greater importance of the labour market in a regional policy setting compared with that in a national policy setting. As far as the labour market itself is concerned, national models concentrate much more on the function of wages as an equilibrating mechanism, whereas regional models concentrate more on quantity adjustments. This reflects the fact that interregional wage differences are smaller as well as more stable than international wage differences, whereas (geographical) mobility is much more important at the regional level. In national models, for example, international migration effects are usually considered to be exogenous (Bodkin, Klein and Marwah, 1991).