Much has been made of the apparent rise in the savings ratio in the December quarter national accounts.
Many commentators and the Federal Opposition have all taken it to mean that the cash splash (AKA the economic stimulus package) failed. They say this without mentioning the rise in retail sales in the month of December, or if they do, they dismiss it.
They have treated it as a hard and fast figure, like so many of the other figures in the accounts.
Far from it: it’s probably the most “rubbery” figure in the accounts. As the ABS warns: “Household saving cannot be directly measured.”
That caveat has been widely ignored, but commentators accepted the rise in the savings ratio to 8.5 as hard fact, without mentioning the ABS’s heavy qualifications, or acknowledging that it is subject to frequent revision, up and down!
In fact this particular figure is one of the most heavily qualified in the national accounts. Here’s what the ABS said yesterday in the release:
The Household saving ratio was 6.6 in trend terms and 8.5 in seasonally adjusted terms in the December quarter 2008. This seasonally adjusted number is the highest since September 1990.
Household saving is not measured directly. It is calculated as a residual item by deducting Household final consumption expenditure from Household net disposable income. As the difference between the two aggregates is relatively small, caution should be exercised in interpreting the Household saving ratio in recent years, because major components of household income and expenditure may be subject to significant revisions. The impact of these revisions on the saving ratio can cause changes in the direction of the trend.
And in explanatory documents linked to the accounts, the ABS had this to say:
Get Crikey FREE to your inbox every weekday morning with the Crikey Worm.
Household saving cannot be directly measured. Rather it is calculated as a residual item by deducting Household final consumption expenditure from Household disposable income as presented in the Household income account. The scope of the Household sector in the National Accounts, and hence the scope of the Household saving ratio includes households, unincorporated enterprises and nonprofit institutions serving households. Due to data limitations it is not possible to completely separate the different sectors.
In other words, the savings ratio is imprecise, hard to work out, and subject to frequent revision.
And a quick look at the current and previous national accounts reveals the treacheries of revision. The September 2008 savings ratio was a high 3.9. At that stage it was the highest for around seven years or so. But by December’s national accounts, it had been trimmed to 3.4 through revisions.
And the March 2007 quarter’s saving ratio was -0.4. The savings figure for the June, 2007 quarter was 0.1; weak and worse said commentators at the time.
By the December 2008 national accounts, that negative reading for March 2007 had become a solid (by recent standards) positive figure of 2.2. And the weak positive for the June quarter had become an even stronger 3.3 in yesterday’s accounts.
And do you think any of those commentators who tut-tutted about our weak savings ratios and culture at the time put their hands up and noted the revision in their latest commentaries?
And finally, the question of the December stimulus package will again rear its ugly head this quarter, and next, and probably for another year as the ABS tries to work out how the money was spent, where it was spent (which industries), and how much was saved.
The December stimulus package featured around $8.4 billion in cash payments to consumers, with the rest spent on first home buyers which runs until June 30 this year. The ABS will now spend the next few quarters analysing its business surveys and other stats to get a better handle on it … and on the impact of the new stimulus package over the rest of this year. In fact, there will be a lot of statistical noise in the retail sales and national accounts for the next four quarters at least.
It will pay to look at the revisions in a year to 18 months’ time.