Research Discussion Paper – RDP 26 Investment Time Lags

Introduction

The lags in the investment process may be divided into two stages. There is one lag between changes in the determinants of the desired capital stock and the decision to invest, and another between the decision to invest and investment expenditures. Empirical studies of investment have frequently estimated these lags together, e.g. [7] [8] and as a result have been able to deduce little about the plausibility of the implied expectations and implementation lags. An alternative approach, e.g. [4], has been to specify the lag distributions separately, but collinearity problems are such that it is necessary to impose prior beliefs about the nature of the implementation lag in order to be able to distinguish between different expectations hypotheses.

This paper uses the building statistics provided by the Bureau of Census and Statistics [2] to investigate implementation lags in Australia.