The trend unemployment rate in Canada: Searching for the unobservable

In this paper, we assess several methods that have been used to measure the Canadian trend unemployment rate (TUR). We also consider improvements and extensions to some existing methods. The assessment is based on four criteria: (i) the extent to which methods provide explanations for changes in trend unemployment; (ii) whether revisions to unemployment gap (UGAP, the difference between the actual unemployment rate and TUR) estimates are well behaved; (iii) if UGAPs provide information about future inflation; and (iv) if UGAPs help explain historical data about wages and consumer price inflation. In our assessment of conformity to the second and third criteria, we use real-time data, i.e., the data available to policymakers at the time of making decisions. We find that while all methods we consider have both strengths and weaknesses, those based on variables thought to determine TUR provide better interpretation and tend to do at least as well as others against the other criteria. These are most promising for future work. Nevertheless, there is considerable uncertainty about the value of TUR, which suggests it would be prudent to use a range of models in research or policy work. While estimates of TUR have declined since the mid-1990s, it is assessed to range between 5.6 and 6.7 per cent in 2018Q4.