Innovative Activity Over the Buisness Cycle:an Alternative Interpretation.

Geroski and Walters (1995) conclude that changes in manufacturing output Granger caused changes in innovative activity and that the latter fluctuated pro-cyclically during the period 1948-83 in the UK. This note shows that their modelling method is inappropriate and results misleading. Alternative VAR models are estimated showing no evidence supporting any of their conclusions. Moreover,neither of the two cointegration relationships (reluctantly) identified by Geroski and Walters are verified.