Abstract

This paper examines spill-over effects among regions. Estimating error correction models of British rates of unemployment and conducting Granger-causality tests, it is found that, on a short-run basis, the West Midlands is a leading territory within a closely knit group of surrounding regions. The paper draws a distinction between a core and a leading region, suggesting the core region, the South East, is a follower, responding to, and magnifying, economic fluctuations. The policy implications are that current regional delineations may not be consistent with the demands of economic coherence at the national level and so is not optimal for the delivery of decentralized spatial economic policy. Indeed, as economic engineering in the high unemployment periphery has consequences for more prosperous areas, it may be better not to allow local administrators too much autonomy as the benefits of a better economic focus may be out-weighed by destabilizing tendencies that decentralization may impose on the whole system. This problem has consequences for, and may be compounded by, policy initiatives at the European level