Prediction Using Panel Data Regression with Spatial Random Effects

Abstract

This article considers some of the issues and difficulties relating to the use of spatial panel data regression in prediction, illustrated by the effects of mass immigration on wages and income levels in local authority areas of Great Britain. Motivated by contemporary urban economics theory, and using recent advances in spatial econometrics, the panel regression has wages dependent on employment density and the efficiency of the labor force. There are two types of spatial interaction, a spatial lag of wages and an autoregressive process for error components. The estimates suggest that increased employment densities will increase wage levels, but wages may fall if migrants are underqualified. This uncertainty highlights the fact that ex ante forecasting should be used with great caution as a basis for policy decisions.