Premium
Seasonal adjustment in the eighties: Some problems and solutions
Author(s) -
Dagum Estela Bee,
Huot Guy,
Morry Marietta
Publication year - 1988
Publication title -
canadian journal of statistics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.804
H-Index - 51
eISSN - 1708-945X
pISSN - 0319-5724
DOI - 10.2307/3315220
Subject(s) - seasonal adjustment , consistency (knowledge bases) , autoregressive integrated moving average , seasonality , series (stratigraphy) , econometrics , smoothing , data series , environmental science , climatology , statistics , geography , time series , economics , mathematics , geology , variable (mathematics) , paleontology , mathematical analysis , geometry
Statistics Canada, like most statistical bureaus, is concerned with problems affecting the accuracy, the consistency, and the usefulness of seasonally adjusted series. The most serious problems affecting the accuracy of seasonally adjusted data in the eighties are related to changes in trading‐day variations caused by changes in consumers' patterns and to the effect of the early arrival of Easter in March (1986) and the beginning of April (1980, 1983, 1985, 1988). The problems affecting consistency mainly result from the lack of uniform practices for the seasonal adjustment of related series and economic aggregates. This lack of uniformity is present in (1) the seasonal adjustment of current observations, which can be done by applying the X‐11‐ARIMA program in four different modes; and (2) the seasonal adjustment of key economic and social indicators, which are aggregates and give different results if adjusted directly or indirectly (through the aggregation of each seasonally adjusted component). Finally, the problem of usefulness of seasonally adjusted data manifests itself for series that are largely irregular, posing then the need for further smoothing.