An Easy Econometric Way of Constructing Input-Output Tables

  • J. F. Divay
  • F. Meunier
Conference paper
Part of the Lecture Notes in Economics and Mathematical Systems book series (LNE, volume 292)


The idea underlying the method is to go directly from micro-economic data of firms over to branch macroeconomic coefficients. When all firms are “pure”, that is to say when each of them produces a single commodity, this is not difficult to do since their mere aggregation by commodity gives the different branchs. Generally, firms are diversified and accounting records only furnish inputs and outputs relating to them, without detailing how each of the inputs is allocated to each output.


Ferrous Metal Order Moment Algebraic Method Generalize Little Square Operating Account 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    This result only holds true when there is no constraint on the coefficients other than that given by the accounting constraint. Thus, in a more general model with constraints on the coefficients, we must reintroduce the equilibrium constraint into the operating accounts of the branches. The balance of operating accounts by branch does not imply that the inputs estimated by the model on all the firms are equal to the actual amounts of inputs observed. If we want to allow for such a constraint, it becomes necessary to restore the balance of the coefficient table by use of RAS-type procedures.Google Scholar
  2. 2.
    In this formula, the technical coefficients relating to the last input are obtained according to (14) and added as the ultimate row in the input-output table.Google Scholar
  3. 3.
    Let us denote as C. the second order moment linking output j to the input; (math) the squared moment. As the model does not contain any constant term, it is not possible to reason directly in terms of non-partial correlation coefficients.Google Scholar
  4. 4.
    This line of reasoning does make it possible to justify the application of the method to firms belonging to diversified industries. When industries continuation are disconnected the model is separable or quasi-separable. For industries whose activities are closest one to another the argument no longer holds.Google Scholar
  5. It appears that an error would be made, when estimating the technology of a branch. If only firms of the industry in that branch were taken, the commodity technology would, thus, be left out when the commodity is used by industries in which it is not the main production. The advantage of the method lies in the aggregation of various activities of firms.Google Scholar
  6. 5.
    In a one-output model, there can be no negative coefficient.Google Scholar
  7. 6.
    Along these lines, one might then think of incorporating as an observation in the allocation model of on input only those firms which use that input in a quantity differing from zero. But this solution — which has been tested — is bad: while it does not change (math), it generally reduces Y’Y and then raises the order of the variance(Y’Y)−1. This result is natural: the firms which produce a given output while not using the input supply a piece of information referring to the exclusivity of the input and the output. They cannot, therefore, be rejected without prejudicing the estimator’s accuracy.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1985

Authors and Affiliations

  • J. F. Divay
    • 1
  • F. Meunier
    • 1
  1. 1.National Institute of Statistics and Economic ResearchParisFrance

Personalised recommendations