'We did not wipe out half the economy': SA's grim GDP data explained

  • While the fall was grim, SA's economy did not halve between the first and second quarters of the year.
  • Stats SA uses annualised data - a technique also employed in the USA and Canada - to headline GDP change between quarters. 
  • But one-off events like SA's hard lockdown can cause the data to be less helpful when extrapolated forward. 

When Stats SA published the country's second quarter growth earlier in the week, the fall in GDP was staggering. 

"Real gross domestic product (measured by production) decreased by a record 51.0% in the second quarter of 2020 owing to the impact of the Covid-19 lockdown restrictions since the end of March 2020," the statistics agency said in its report.

A 51% fall would mean the economy shrank by half between the first and second quarters of the year, a drop bigger than the heights of the Great Depression. 

But a footnote changes the meaning of the -51% drop. And no, half of the economy did not evaporate due to the hard lockdown. "We did not wipe out half the economy," noted University of the Free State Professor Philippe Burger in a statement.  

The footnote states that Stats SA uses annualised data for its GDP predictions. This means the GDP change is shown as if it continued and compounded at the same rate for a year, rather than a simple percentage change between two quarters.

Why? Because it makes comparison between different time periods easier. 

When the annualisation is removed, the actual simple percentage change between the two quarters is -16.4%. While still grim, this is not as bad as half of SA's GDP vanishing.

Among economists and data wranglers in South Africa, contention is now brewing about which measure to use. While the annualised approach has been used for years, some argue that during the pandemic it creates unnecessary confusion.

The reason is that the hard lockdown - which shuttered much of the economy in April and May - is not something that would likely be repeated again and gain over the next three quarters. But annualisation is calculated on the basis that it would be. It's an extrapolation forward, and assumes that each quarter would see an additional -16.4% decline.  

"The -51% quarter-on-quarter seasonally adjusted annualised basis refers to quarter-on-quarter growth for quarter two of 2020 that is annualised to see what the outcome for the year would be if each of the four quarters of 2020 saw the same suppressed level of economic activity that the second quarter of this year did," noted Annabel Bishop of Investec. 

When one-off events like a hard Covid-19 lockdown occur, causing large fluctuations in quarterly GDP, annualisation starts to become less helpful.  

In September 2020, for example, gyms, hairdressers, restaurants and cafes are again open. In April, by contrast, you couldn't buy a hot chicken at a supermarket.  

South Africa has already moved to lockdown Level 2, and President Cyril Ramaphosa was signalling that a further relaxation of lockdown restrictions may be on the cards. 

So what is actually going on?

Stats SA executive manager of short-term indicators, Michael Manamela, said Stats SA had always reported the annualised GDP as the headline number. 

"This practice is the case for South Africa and a few countries such as the USA and Canada. Although the seasonally adjusted annualised rate is referenced as the headline, Stats SA has always provided estimates for the year-on-year estimates as well as data to derive the not annualised quarter-on-quarter estimates," said Manamela.

On Tuesday, this was the case, with Stats SA giving three figures in its presentation - the annual rate (-17.1%); the quarter-on-quarter not annualised rate (-16.4) and the quarter-on-quarter annualised rate (-51%). 

"Historically, South Africa has used the seasonally adjusted annualised quarterly GDP as the headline number. Globally there is no one standard, but many countries' headline number is either the year-on-year or the seasonably adjusted quarter-on-quarter not annualised," Manamela said.

Is this a problem only in South Africa?

This was not a conversation only taking place in South Africa.

Economic analyst at Tutwa Consulting, Heinrich Krogman, notes that the same discussion happened in the US recently when they announced their GDP figures.

The New York Times reported that second quarter GDP figures in the US could be looked at in different ways - either as a 35% annualised drop or a 10% quarter-to-quarter non-annualised drop.The newspaper chose to lead with the non-annualised number, noting that when annual rates were applied to short-term changes, they could be "misleading". 

But the annualised GDP figures are useful for comparisons, noted Krogman. 

"Reporting data (or GDP figures for that matter) in this format helps in making a quick comparison regardless of the time period," said Krogman.

What impact does it have?

Investment strategist at Old Mutual Multi-Managers, Izak Odendaal, said while the manner in which the data is presented does a lot for how it is perceived, the severity of the GDP contraction announced on Tuesday should not be lost on South Africans.

On the other hand, Odendaal said, the modern world was experiencing economic realities the likes of which had not been seen before and this had to be considered when considering GDP data and how it was presented.

"Normally annualising isn't an issue but this wasn't a normal quarter. Neither is wrong, as long as people understand what the numbers mean. We've always published the annualised number in SA, but some countries don't. As long as you compare countries on the same basis," Odendaal said.

We live in a world where facts and fiction get blurred
In times of uncertainty you need journalism you can trust. For only R75 per month, you have access to a world of in-depth analyses, investigative journalism, top opinions and a range of features. Journalism strengthens democracy. Invest in the future today.
Subscribe to News24
Brent Crude
All Share
Top 40
Financial 15
Industrial 25
Resource 10
All JSE data delayed by at least 15 minutes morningstar logo
Company Snapshot
Voting Booth
Please select an option Oops! Something went wrong, please try again later.
Yes, and I've gotten it.
21% - 384 votes
No, I did not.
52% - 949 votes
My landlord refused
27% - 501 votes