Abstract
There are two groups of models. The one type of model is designed to forecast the immediate future. The other type is concerned with the analysis of structural change and economic growth. The dichotomy goes back to the underlying economic theory. There are not only two groups of models, but there were two groups of economic theories well before there were formal models.
A paper, which is substantially this chapter, less Section 7.5, with an additional introduction, will also appear under the title “Short versus long-term economy models” [27a].
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Notes to Chapter VII
Yet, I understand this is in fact being done by some practitioners. See Klein et al. [33].
See also Goldberger [20], p. 373.
Accordingly, there is a case for a method of simultaneous estimation, which relates the estimates of model-parameters to the fit of the sample-periods cumulative ‘forecasts’.
Yet this was done by Van den Beld [3].
The special case of ‘forecasting’ initial conditions, for which the statistics are not yet available, will arise. See Heesterman [27].
Note, that the model satisfies the causal chain requirement.
A flexible accelerator, without a financial component, would be (Math) where k is the capital-output ratio, and α an adjustment-percentage; entrepreneurs are assumed to adjust capacity to sold output to a fraction of a, which would in this case be 90%.
In fact it is almost standard procedure to formulate short-term forecasting models in terms of percentage increments of the accounting variables. The main reason for this practice is that it allows estimates of the structural relations to be free of heteroscedasticity, as well as of positive serial correlation of the error term. Definition identities must of course be approximated by means of the Taylor expansion. Now if this method is followed, one can simply apply (7.4.3), with γ = 1. Different growth-rates of exogenous variables are now fed in as ‘levels’ of the (changes in) accounting variables. The outcome for the dependent variables will then of course also be in terms of percentage increments; which can be compared with their average increments (trend) over the sample period.
In fact a fixed accelerator-type of investment demand is implied by the combination of capacity being equal to sold output and the use of a capital output ratio. This is not normally a converging adjustment mechanism. (See Section 6.3.)
Rights and permissions
Copyright information
© 1970 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Heesterman, A.R.G. (1970). Dynamic Adjustment Models and Their Convergence. In: Forecasting Models for National Economic Planning. Springer, Dordrecht. https://doi.org/10.1007/978-94-017-6214-4_7
Download citation
DOI: https://doi.org/10.1007/978-94-017-6214-4_7
Publisher Name: Springer, Dordrecht
Print ISBN: 978-94-017-5793-5
Online ISBN: 978-94-017-6214-4
eBook Packages: Springer Book Archive