As the extreme impacts on a raft of economic data collected by them became clear the ABS, in many instances, have suspended the production of their Trend estimates.
The reason the ABS have provided is along the lines of the following (which relates specifically to the Labour Force data but applies to a range of data sets)…
“Trend and seasonally adjusted Labour Force Survey estimates are important for understanding the current state of the labour market, and providing medium and long-term signals.
However, unusual, sudden and/or extreme changes in the labour market (e.g. changes to regular and seasonal patterns in employment, working hours, job search or availability as a result of COVID-19) can be difficult to account for in the short-term in both seasonal adjustment and trend estimation processes.
The Labour Force trend series have been suspended from April 2020, until more certainty emerges in the underlying trend in labour market activity over the COVID-19 period. This is in recognition of the time series shocks in the labour market between the March 2020 and April 2020 data points. While trend estimates are generally the best measure of the underlying behaviour of the labour market, any large changes in the labour market as a result of COVID-19 will likely make it difficult to estimate a reliable short-term indicator of the trend.”
I thought it might be useful to demonstrate what this effect would look like using a particularly distorted and shocked data set; short term visitor arrivals.
If we only include data on short term arrivals up to and including February 2020 the first chart shows the original ABS data (dotted line) along with the ABS Trend series (red line) and our own Conus Trend series (blue line). As you can see, although the ABS and Conus methodology might vary slightly the resulting Trend series were remarkably similar at all times.
However, if we extend the data beyond February (at which point the ABS Trend was suspended) and look at the Conus Trend we can see the kind of distorting effect that this dramatic shift in the original data has had…not just on the Trend since February but for years before then. This particular data set is an extreme example given that arrivals have fallen by some 99.5% since February but hopefully provides some insight into why the ABS have suspended a number of Trend estimates in the current circumstances.