Making meets spending
After a careful handover from the Reserve Bank, Stats SA is ready to go it alone with publishing SA’s official GDP statistics
n a move designed to produce more accurate national accounts, Stats SA has brought the entire measurement of gross domestic product (GDP) under its own roof, with the Reserve Bank’s blessing.
Since the 1940s, the Reserve Bank and Stats SA have split the measurement of GDP between them, with the Bank measuring GDP from the expenditure side of the economy and Stats SA responsible for using the production approach.
SA is the only country in the world which split the measurement of its national accounts in this way, according to Joe de Beer, the deputy director-general of economic statistics at Stats SA.
Economic theory suggests that the two different ways of measuring GDP should yield the same result. In most countries they don’t. In SA the gap between the two approaches — the residual — has been large and mostly growing for many years. This has undermined the credibility of SA’s GDP data.
It made sense to bring the Bank’s expenditure-side responsibility into Stats SA because the latter is better placed in terms of statistical skills and access to granular data to do the job, according to De Beer.
Because of this, he believes Stats SA is in a better position to harmonise demand and supply than the Bank, “so the quality of SA’s GDP estimates will improve”.
In fact, Stats SA says international evidence shows that it is “inefficient and counterproductive” to handle large-scale statistical operations in anything other than
Ia fully integrated environment. The first integrated GDP time series was published last week with historical quarterly revisions going back to 2010. Stats SA has been transparent about its methods and invited critical feedback from users in a bid to keep improving the new integrated approach.
BNP Paribas Securities economist Jeffrey Schultz thinks that SA has taken a step towards producing more accurate national accounts.
For instance, he calculates that the implicit residual in the revised numbers has fallen to only minus R4.6bn on average over the past five years compared with minus R26bn in the Bank’s old residual series.
The fact that there is now much closer alignment between the production and expenditure side GDP estimates is one of the most “important and encouraging” developments, says Schultz.
Stats SA says it will consider it a measure of success if the residual can be reduced to a very small number which averages zero over several years.
There have been some mutterings in the markets about the timing of the first-quarter 2016 GDP release, however, given that the much-anticipated figure will now be published only on June 8, after Standard & Poor’s and Fitch have announced their rating actions on SA.
The number is expected to be a shocker. The Bank’s forecast for first-quarter GDP growth is barely positive, with the mining and agricultural sectors expected to be the biggest drag on the outcome.
Citi Bank is expecting growth to have contracted by 0.7% in the first quarter compared to the final quarter of last year.
But De Beer says: “There should be no [suspicion] in users’ minds that we are open to influence.”
To counter precisely this suggestion, Stats SA invited economists and financial journalists to a series of workshops during 2015 to explain the handover from the Bank, a process which has taken four years and been open to external scrutiny.
At the outset, Stats SA recruited 13 university graduates whom it has trained with the help of a senior International Monetary Fund (IMF) consultant as well as two international experts, Roger Jullion and Ross Harvey, the former respective heads of national accounts at the Canadian and Australian statistical agencies.
Towards the end of 2015, Jullion and Harvey produced two audit reports on the unit’s readiness and suggested areas for improve-