FATHER'S DAY PROXIMITY EFFECT
Father's Day occurs on the first Sunday of September each year. This means that Father's Day may fall on any of the first seven days in September. In 2002, Father's Day fell on the first of September.
If Father's Day expenditure occurred wholly in September any systematic pattern would be catered for in the seasonal adjustment process. However, some of the expenditure occurs in the days leading up to Father's Day and, depending upon how close Father's Day is to the start of September, the expenditure will move between August and September. The previous seasonal adjustment process did not adequately cater for these instances which are referred to as the Father's Day proximity effect.
In 1996, the ABS investigated the Father's Day proximity effect. For more information refer to the feature article in Australian Economic Indicators, December 1996 (cat. no 1350.0). At that time, the evidence for a Father's Day effect was not conclusive. However, recent investigations have used a regression model that is embedded in the seasonal adjustment process to correct the Father's Day impact if it is proved statistically significant.
The regression method uses the irregular values from August and September obtained from the seasonal adjustment program (X11) for each series and, if the irregular values in the affected years deviate significantly, this deviation is used to estimate the Father's Day proximity effect. This methodology has also been recently implemented to address the Easter proximity effect (refer to the appendix in the July 2001 issue of this publication).
IMPACT ON RETAIL TRADE
The Father's Day proximity effect was found to have a significant impact in one main Retail industry; Department stores. Other significant impacts occurred in other industries; for example, Pharmaceutical, cosmetic and toiletry retailing, Hotels and licensed clubs, Other food retailing, and Supermarkets and grocery stores. Significant impacts were found in the New South Wales total state series, and the total Australian Retail and Hospitality/Services series. The correction for this effect has been included in the seasonal adjustment process from this year's seasonal reanalysis and has led to historical revisions to the affected series, especially for August and September months.
The correction for the Father's Day proximity effect reduced the August 2002 seasonally adjusted estimates for Australian Retail and Hospitality/Services by 0.8 percentage points and Department stores by 3.4 percentage points.