Estimating the NAIRU : an analysis of the Bank of Canada model
Date
1991
Authors
Paulsdorff, Ulrich Gustav Fritz
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The Bank of Canada proposes a model consisting of four equations to estimate simultaneously the NAIRU, the nonaccelerating-inflation rate of unemployment. The model consists of an aggregate production function of the kind used in neo-classical growth models, an Okun's Law equation, a Phillips curve and a fourth equation that defines the NAIRU to be determined by Unemployment Insurance payments and demographic factors. This last equation is substituted into the system wherever the NAIRU appears. The Phillips curve as the centrepiece of the model is of the accelerationist type. The theoretical concepts underlying the Phillips curve and the production function lead to certain restrictions being imposed on the parameters of the model.
The importance of having robust and reliable estimates of the NAIRU and the controversial theoretical basis of the model motivate the analysis in this thesis of the restrictions and the estimates of the model. Less restricted model versions are specified and likelihood-ratio tests are used in order to test the restrictions. For the analysis of the estimates of the Bank of Canada model and of the less restricted versions of the Bank of Canada model estimated here, expected values are determined on the basis of theoretical considerations and previous empirical studies. In addition, NAIRUs are generated with the estimates of the different models to see whether they are meaningful. The restrictions are mainly rejected but the less restricted models are not satisfactory in comparison to the Bank of Canada model because their parameter estimates often have the wrong sign or are out of the expected range. Furthermore, the NAIRUs generated have unrealistic values in many cases. On balance, the results fail to support the theory underlying the model. This does not necessarily mean that the theory is wrong. It can very well be the case that the model does not appropriately reflect the theory. Nevertheless, the NAIRU estimates of the Bank of Canada model are not reliable and should be regarded with great care.