[TOP STORY] Making sense of the GDP data

SIMON BROWN: I’m chatting now with Joe de Beer, deputy director-general responsible for economic stats at Stats SA (Statistics SA). Joe, I appreciate the early morning. Before we jump into some of the details, a bit of background. Obviously these are data points rather than physical points, but are there challenges or have there been over the last two years re Covid and collating, collecting and computing the GDP data?

JOE DE BEER: Good morning, Simon. Thanks for having us. Yes, indeed. The Covid period has been quite challenging. We had to change our business model quite a bit, economic statistics especially were mostly head office-based and my staff was running telephone surveys, emails – and then Covid struck. We had to find a way to get everybody to work from home. So we had a few hundred people with desktop PCs who had go home, provide data for internet access. Once we got that sorted, the second complication was that many of our respondents also started working from home.

SIMON BROWN: Oh, yeah.

JOE DE BEER: The landline numbers we had then were no longer being answered, because people were maybe using their cellphones to work from home. So it took us a while to get up to speed. You might remember that when Covid  struck, most of our monthly releases were delayed by a  couple of weeks. But we managed to catch up over time.

On the social stats side with households, the complication was we couldn’t send the field worker to a household to interview them, unless that was of course a labour court……1:41 survey because of the Covid restrictions, because the households wouldn’t want to receive our field workers, and we also couldn’t send a field worker to somebody whose health status we didn’t know. So we had to change that to a telephone survey rather than a face-to-face one, which had implications on the sample sizes and lots of other hassles that we actually solved. But we are sort of back on track.

SIMON BROWN: Okay. We are sort of exiting the pandemic, and we were all wiser and better for it.

The data out yesterday – [GDP] Q4 up 1.2%, the annual rate for 2021 is 4.9% growth. That does, however, still leave us below pre-pandemic levels in that our economy is smaller than it was in December 2019. That to me was one of the key points.

JOE DE BEER: Yeah. That’s very true. But I think sometimes GDP takes a longer-term view, especially if we talk about the annual numbers. Even for the couple of years we had pre-pandemic the growth rates were very muted. I think it was a long time coming. It’s very difficult for a big thing like the economy to suddenly just kick start and start shooting the lights out. So in my view, although we don’t do any forecasts or ……2:53 – I leave that to my colleagues in the policy department – I think it’ll have to take a gradual uptick in the economy and it should be spread across all industries.

If you look at the quarter numbers we released now, it was not growth that we saw everywhere. Half of the industries that we distinguish contributed about a third of a point each, and half of them detracted about 0.1 of a point. So the growth is not yet evenly spread across the economy and, until that happens, it shows us you won’t have sustained high-growth levels. Everybody needs to be able to chip in.

SIMON BROWN: Yeah. I like your point there that an economy is slow and we take a long-term view. This isn’t a racing car, you don’t turn it on a dime. Those big industries – mining, agriculture, manufacturing – were the ones; and certainly in mining and agriculture I think we see that. We talked to miners and farmers and they’ve had a really booming couple of years. Manufacturing is coming in. We are seeing some of those sectors having really strong growth.

JOE DE BEER: What makes the annual number quite interesting is that we must still remember that we are comparing 2021 with 2020, and 2020 was an abnormal year. You have so many base effects at play because in 2020 you had a negative 6.4%. So any number for 2021 would probably be very [strong] growth, purely because you’re comparing it to such a low number. In a way the same holds for the fourth quarter, because now in the fourth quarter we have 1.2%. If we think about the old days, that would be nearly 5% if we annualised the numbers – and that’s very strong growth. But it comes from that negative number again.

So for both the quarter and the annual we need to take into account what the base effects are and what is really the underlying growth.

SIMON BROWN: The largest sector, our largest industry in our country, is finance, followed by personal services. We kind of think of ourselves as a mining country – and we are. Mining is significant, so is agriculture, but actually finance is our biggest industry by a way.

JOE DE BEER: Yeah. Let’s just also be clear, although we call it finance, it’s still an abbreviation for ‘finance, real estate and business services’. So it’s ……5:09 really business itself that’s included in there.

Personal services, as you rightly point out, is the second-largest one. That also has to do a little bit with the reclassification we did last year, when we benchmarked the GDP. We moved all the health education from what was previously called ‘government and community services’ into ‘personal services’, which is a little more accurate to do. So within the  personal services, the strong growth that we saw in the fourth quarter was Covid-related because it was expenditure on health. That was the main runner …… 5:39 for the fourth quarter.

SIMON BROWN: A last question. You made some tweaks to the third quarter, and this is fairly standard. The numbers which you released yesterday – we are what, 10 weeks from the period end – part of the process I imagine is you go back and check on the previous quarter and there might be some changes. But do you now kind of lock that in, would you revisit that Q3 again, or is there a sense of we are happy with that now, we have done the tweak and we think that we’ve got it as accurate as we can [get it]?

JOE DE BEER: No, it’s not yet locked. The reason we divided the Q3 now was because we had revised data, especially on agriculture, from the Department of Rural Development and Land Reform, where we get our agriculture data. They have  quite significant changes to the data. And then even for some of our own surveys, although we base-record the GDP on monthly surveys, it takes us maybe a year, a year-and-a-half to start getting more accurate annual surveys on board, and those annual surveys have much more detailed financial information that gives us an opportunity to make a more reliable estimate of the GDP.

So these quarterly numbers are sort of kept open for revision for quite a long period.

SIMON BROWN: Okay. The changes are typically – I see this coming out of the US data and in other markets as well – those changes are usually small.  I don’t want to say they’re insignificant, but they really are smaller. They’re the rounding errors, really.

JOE DE BEER: Yeah. The saying in our circles is that if you don’t revise, something is wrong, because over time you need to get more accurate and more reliable data. So if you are able to estimate something big, like the economy within 12 weeks, and you are never going to change it, well, then you should maybe revisit your methods.

SIMON BROWN: I take that, and I really like that point. Economies are slow. They don’t turn on a dime; it really is about the longer term. Joe de Beer, DDG responsible for economic statistics at Stats SA, I really appreciate the time this morning.