Seasonal adjustment is proven to be a useful tool for economic analysis. However, as any other processes, it has advantages and disadvantages. To gain from the former and diminish the impact of the latter, one expects a proper level of quality in the seasonally adjusted data. Measuring that level offers challenges from the initial definition of quality to the practical day-to-day application of the process. The authors share Statistic Canada's experience in that regard from both a managerial point of view and a very operational approach. Best practices such as horizontal review and quality guidelines will be discussed as well as real-life operational challenges such as mass-production, varying reporting and dissemination frequencies and abrupt seasonal changes.
Keywords: Seasonal adjustment; Best practice; Quality
Biography: Susie Fortier is the chief of the Time Series Research and Analysis Centre at Statistics Canada. She teaches various time series courses in English and in French, at the institute's Training Centre.
Her research interests focus on time series reconciliation and benchmarking, seasonal adjustment, price indices methodology and SAS programming. Ms. Fortier is also the production manager of the Survey Methodology Journal.