Summary
The seasonal adjustment package used by Statistics New Zealand, X12-ARIMA, has an automatic procedure for dealing with outliers (observations which are far removed from the others in the series), which works well in most cases. However, in certain circumstances outliers need to be dealt with explicitly.
In this paper we consider the situation of the Quarterly Gross Domestic Product series, in which there are outliers in two consecutive March quarters. When the second of these outliers was the last point in the series the automatic procedure treated neither quarter as an outlier. A simple, practical solution was to treat one of the outliers explicitly by prior-adjusting the original series. This enabled the other outlier to be identified automatically by X12-ARIMA in the usual way.