The NTV Model for Total Factor Productivity

Andrew Smithers
World Economics • Vol. 20 • No. 2 • April–June 2019, 1st July 2019


This paper revises and extends my previous one on TFP * in which I proposed that the NTV model should be preferred to those which follow the current consensus approach. I include in this paper the equations for calculating TFP by two different approaches based on completely independent data and show that these can be used to test the validity of the model.

Both the NTV and the consensus models seek to determine the contribution made by changes in technology to total growth. The growth rate determined by changing technology is termed TFP, with the balance coming from growth in the labour supply and the volume of the capital stock. Both models assume that the rate of change in the capital stock depends on the level of investment net of capital consumption.

The differences between the two models include the definition of capital consumption and assumptions about corporate and investor behaviour. The results of the models differ significantly and have important consequences for economic policy. In setting out these differences I explain my reasons for preferring those used in the NTV model.


* Building a Testable Model to Estimate Total Factor Productivity World Economics Vol. 18 No. 2 April-June 2017.

Please click on the Download link below for the pdf of the complete paper

This paper and other papers are also available on World Economics website, once you have registered.
The NTV Model for Total Factor Productivity