|
Abstract:
|
Early in the COVID epidemic Statistics New Zealand (StatsNZ) recognised that there was a strong possibility that COVID would have significant effects on its time series. StatsNZ produces seasonally adjusted and trend estimates for many thousands of series, so the initial decision was to cease publishing these outputs, while continuing to create and internally monitor the outputs closely. Over the last two years we have found for many series COVID, while affecting the data, has not materially changed their overall behaviour. Thus the outputs could be published with minimal modifications of existing methods. However, some do not appear to have stabilised, so the question remains, how can we recognise when they do stabilise and then, are the component's behaviour similar or dissimilar to their behaviour prior to COVID?
|
