In year t-1
Figure 18.8 Energy intensities in the Netherlands, 1970-96, in toe/1000 US$ (1990)
made if the nature of the shocks is not understood. (A shock can be defined as an influence on one variable in an estimated relationship which is not explained by the other variables. As a shock is exogenous to a certain specified model, inclusion of other variables may endogenize such a shock and this certainly will improve predictions. Imagine, however, how difficult it must have been to predict the oil shock back in 1971.) This point was discussed by Labson and Crompton (1993) but has remained largely ignored in the literature, probably because it deals with more sophisticated econometric techniques such as cointegration.
Was this article helpful?