A few weeks ago the Australian Bureau of Statistics (ABS) published the June quarter 2021 Australian National Accounts: National Income, Expenditure and Product (Cat. No. 5206), which means we can update the Practical Economics quarterly Australian State and Territory Gross State Product (GSP) series.
We’ve been very busy on our day job so not only are we a bit later than usual, but this post is also going to be brief and our quality control will sink to a new low.
An additional complication in our code revealed itself in our June quarter calculation, which required a bit of re-writing. It doesn’t affect our previous results and is specific to the June quarter of each year.
Also, we have a full year, or four quarters, of indicators without an annual benchmark, which will persist until the 2020/21 State Accounts are released by the ABS on 19 November 2021[1]Soon after which we get the next indicator.. This means that the indicators are as unconstrained from the benchmark series as they are going to get. That is, at their most unreliable.
We always encourage you to read our disclaimers here and here.
Anyway, to a quick tour of the results. Here are the June quarter 2021 quarterly and annual Practical Economics GSP results.
June Quarter 2021 Practical Economics Gross State Product, Percentage Changes, Chain Volume Data
| State | Quarterly Growth Dec-20 to Mar-21 | Year-on-year Growth Mar-20 to Mar-21 |
| NSW | 1.65 | 9.56 |
| VIC | 1.74 | 9.85 |
| QLD | 1.1 | 8.71 |
| SA | 0.94 | 10.62 |
| WA | 0.54 | 4.95 |
| TAS | 0.62 | 11.98 |
| NT | 1.07 | 5.98 |
| ACT | 1.96 | 7.89 |
| AUS | 1.34 | 8.78 |
It’s difficult to tell what is really going on in the various state and territory economies given the impact of COVID-19 restrictions.
In the June quarter 2021 New South Wales, Victoria, Queensland and South Australia continued their rapid recoveries from severe downturns in the middle of 2020. In contrast, growth in Western Australia was much slower, but this largely reflects that it did not experience the severity of downturn as did the eastern states.
Lockdowns in response to the current eastern seaboard COVID-19 outbreak began in late June 2021, so the economic impacts will be felt in those States from the September quarter onwards.
Here are our indecipherable 19-year growth charts. The impact of COVID-19 lockdowns in 2020 are the most notable feature in every chart.
To determine whether the economy has returned to its trend before COVID-19 restrictions, we constructed an output gap based on a linear trend of GSP for the five years leading up to the COVID-19 downturn.
We can see Australia is still 1.73 percentage points below its June quarter 2021 potential output, which was a narrower gap than the 3.1 percentage points in the March quarter.
The output gap in the States and Territories varies greatly. NSW, Victoria Queensland and Wester Australia remain below trend output (a positive value in the table represents below potential output).
However, South Australia, Tasmania and the Northern Territory were their above respective output trends in the June quarter 2021.
June Quarter 2021 Output Gap, Percentage Gap, Chain Volume Data (positive gap means output below trend)
| State | Output Gap (%) |
| NSW | 1.75 |
| VIC | 2.55 |
| QLD | 2.65 |
| SA | -0.3 |
| WA | 0.56 |
| TAS | -0.88 |
| NT | -2.4 |
| ACT | 1.76 |
| AUS | 1.73 |
As indicated above we’ve updated our code owing to the fact we didn’t foresee that we would have a situation with a full financial year worth of indicators but no annual benchmark.
This is important because the code has to decide whether the indicator time series exceeds the benchmarks available. It does this so it can extrapolate previous adjustment factors to the unbenchmarkable indicators if necessary.
We implicitly assumed that, if we had a full year’s worth of indicators, we would have a annual benchmarks to cover them. This means we linked the extrapolation decision to whether the indicator series could be divided by four with no remainder.
For example we created a variable ‘remainderLengthOfOriginalSeries’ as the residual of the length of the indicator series divided by four.
remainderLengthOfOriginalSeries=len(national_Quarterly_Industry_GVA)%4
We then used it in a ‘if’ statement for extrapolation purposes.
if remainderLengthOfOriginalSeries!=0:
Consequently we now create a variable ‘state_Annual_Industry_TFI_length’, which is the length of the benchmark series multiplied by four.
state_Annual_Industry_TFI_length=len(pd.read_csv(workingDirectory+’/state_an_cp_NSW.csv’))*4
We then calculate ‘remainderLengthOfOriginalSeries’ as the difference between ‘state_Annual_Industry_TFI_length’ and the indicator series.
remainderLengthOfOriginalSeries=national_Quarterly_Industry_GVA_lenth – state_Annual_Industry_TFI_length
What do you think of our updated series? If you have any comments, suggestions or want to purchase full results then please contact us.