Testing the Long Run Effect of Investment on Output in the Presence of Cointegration.

A major empirical interest is whether a permanent change in economic fundamentals produces a growth effect. However, a direct time series analysis of this hypothesis may not always be feasible due to a lack of such events. This paper explains why a test regarding the long run effect of a temporary change in investment share may, under appropriate conditions, provide indirectly the answer regarding the effect of a (possibly hypothetical) permanent change in investment share.