Testing the Stability of the Canadian Phillips Curve Using Exact Methods

Postulating two different specifications for the Canadian Phillips curve (a purely backwardlooking model, and a partly backward-, partly forward-looking model), the authors test for structural breaks in the parameters of the equation. In each case, they account for the possibilities that: (i) breaks can be discrete, or continuous, and (ii) available data samples may be too small to justify using asymptotically valid structural-change tests. Thus, the authors use recent testing procedures that are valid in finite samples, applying the Dufour-Kiviet (1996) methodology for discrete-type breaks, and the Dufour (2002) Maximized Monte Carlo test method for continuous-type shifts. The second test accounts for nuisance parameters that appear only under the alternative. The proposed alternative is a Kalman-filter-based time-varying-parameter specification, with coefficients that follow random walks. The authors find evidence for linear and non-linear breaks, the latter being characterized by continuous and unpredictable-type shifts in the inflation-dynamics coefficients.