TESTING THE NEG MODEL : FURTHER EVIDENCE FROM PANEL DATA

Local wage variations in the UK are explained by two non-nested rival hypotheses. The first derives from new economic geography theory, in which wages depend on market access. The second come from urban economics theory, giving a reduced form with wage rates dependent on employment density. The paper examines whether one of these rivals is encompassed by the other by fitting an artificial nesting model using three alternative panel data estimators. The estimates indicate that neither hypothesis is encompassed by its rival, suggesting a need for new, more comprehensive, theory.