WEBVTT 00:00:05.099 --> 00:00:08.599 Hello and welcome to the So What podcast, in 00:00:08.599 --> 00:00:11.099 which political economic analyst J .P. c 00:00:11.099 --> 00:00:13.919 discusses the issues uppermost in the minds of 00:00:13.919 --> 00:00:16.640 South Africans. You can find a written version 00:00:16.640 --> 00:00:19.679 of this content on J .P.'s website, jplandman 00:00:19.679 --> 00:00:24.120 .co .za. I am Ruda Landman, and I am your host. 00:00:26.359 --> 00:00:30.620 This podcast accompanies J .P.'s newsletter dated 00:00:30.620 --> 00:00:35.100 the 21st of July, 2025. under the title Getting 00:00:35.100 --> 00:00:39.979 a Grip on the Informal Sector. The outgoing Capitec 00:00:39.979 --> 00:00:43.880 CEO, Harry Fourie, generated quite a debate recently 00:00:43.880 --> 00:00:48.020 on South Africa's informal sector and the unemployment 00:00:48.020 --> 00:00:51.820 rate. And this is in response to that, right 00:00:51.820 --> 00:00:54.500 JP? Yes. Where did you go to find information? 00:00:54.960 --> 00:00:57.539 Well, I paid a visit. I went to see Joe de Beer, 00:00:57.700 --> 00:01:00.119 who's the Deputy Director General at Stats SA, 00:01:00.399 --> 00:01:04.319 responsible for economic statistics. And he was 00:01:04.319 --> 00:01:07.379 very helpful, very generous with his time and 00:01:07.379 --> 00:01:09.239 experience. And he's clearly a guy who's got 00:01:09.239 --> 00:01:11.879 his feet very, very firmly on the ground. He 00:01:11.879 --> 00:01:14.599 knows what's going on. I also relied on reports 00:01:14.599 --> 00:01:17.620 by the Bureau for Economic Research at Stellenbosch 00:01:17.620 --> 00:01:21.019 and also on work that was done by Professor Johan 00:01:21.019 --> 00:01:23.739 Verrie from the Stellenbosch Economics Faculty. 00:01:23.980 --> 00:01:26.480 So those three sources helped me to try and come 00:01:26.480 --> 00:01:30.060 to grips with this rather complex topic. Okay, 00:01:30.120 --> 00:01:34.260 so let's start with the basics. How is GDP determined? 00:01:35.019 --> 00:01:39.700 That is indeed the starting point. We follow 00:01:39.700 --> 00:01:43.040 in South Africa the international standard of 00:01:43.040 --> 00:01:46.980 SNA, and SNA is simply a system of national accounts. 00:01:47.219 --> 00:01:50.659 It is run by the United Nations. It was started 00:01:50.659 --> 00:01:55.599 in 1947. The first one was issued in 1953, and 00:01:55.599 --> 00:01:58.219 there are regular updates, and the latest update 00:01:58.219 --> 00:02:02.120 was in 2025. So what we can see is that there 00:02:02.120 --> 00:02:04.840 are many, many people in many countries in the 00:02:04.840 --> 00:02:08.180 world over many decades who are seized with this 00:02:08.180 --> 00:02:11.800 matter. And South Africa obviously is one of 00:02:11.800 --> 00:02:16.699 them. If you say it was published, what was published? 00:02:16.879 --> 00:02:19.500 Is it a kind of template? It's a template. It's 00:02:19.500 --> 00:02:23.400 a template with rules and standard formulations 00:02:23.400 --> 00:02:26.400 that must be used. So it's really a globally 00:02:26.400 --> 00:02:29.060 synchronized and globally standardized system. 00:02:29.710 --> 00:02:33.610 And South Africa applies that. Now, what happens 00:02:33.610 --> 00:02:37.050 is that the S &A, the System of National Accounts, 00:02:37.050 --> 00:02:40.370 measures the transactions that take place in 00:02:40.370 --> 00:02:43.949 the economy through three accounts, a production 00:02:43.949 --> 00:02:46.710 account, an expenditure account, and an income 00:02:46.710 --> 00:02:49.389 account. And basically what happens is you trace 00:02:49.389 --> 00:02:52.509 the flow of money as money moves through the 00:02:52.509 --> 00:02:54.710 economy, and you capture it in the different 00:02:54.710 --> 00:02:57.389 accounts. And that gives you a bird's eye view 00:02:57.389 --> 00:03:00.110 of what's going on in the economy. That bird's 00:03:00.110 --> 00:03:02.770 eye view is what we call GDP or gross domestic 00:03:02.770 --> 00:03:06.009 product, which all our listeners will be familiar 00:03:06.009 --> 00:03:09.210 with. Now, it is very important to note that 00:03:09.210 --> 00:03:13.490 gross domestic product is the value added by 00:03:13.490 --> 00:03:16.110 all economic actors. So whether you're a household, 00:03:16.310 --> 00:03:20.449 individual, a podcast studio, a bank, a mine. 00:03:21.080 --> 00:03:23.719 They're all economic actors. And the issue is 00:03:23.719 --> 00:03:27.219 how much value do they add onto the economy? 00:03:27.620 --> 00:03:31.580 Why is this important? It is only about the value 00:03:31.580 --> 00:03:35.680 added. You have to deduct from what an economic 00:03:35.680 --> 00:03:38.500 player produces, a bank or a podcast studio. 00:03:38.759 --> 00:03:41.939 Whatever they produce, you have to deduct the 00:03:41.939 --> 00:03:44.860 input cost. What does it cost them to produce 00:03:44.860 --> 00:03:47.300 that good or service? If you don't do that, you 00:03:47.300 --> 00:03:50.419 do double counting. And you don't want to do 00:03:50.419 --> 00:03:53.400 that. It'll give you a completely wrong reading. 00:03:53.900 --> 00:03:57.699 And many of the Qazi economics people, if I can 00:03:57.699 --> 00:04:00.939 call it that, make this basic mistake of not 00:04:00.939 --> 00:04:05.280 deducting the input costs. And then you sit with 00:04:05.280 --> 00:04:09.199 unreliable data. Unpack it in terms of, say, 00:04:09.419 --> 00:04:13.400 someone who sells tomatoes on the street. OK, 00:04:13.439 --> 00:04:17.579 let's take a street vending lady selling on the 00:04:17.579 --> 00:04:20.779 street tomatoes and sweets. Let's say that she 00:04:20.779 --> 00:04:24.319 makes 2000 rand a day. That is what she takes. 00:04:24.540 --> 00:04:26.939 That's her total income. And now, of course, 00:04:27.019 --> 00:04:29.259 she's registered nowhere. So she doesn't appear 00:04:29.259 --> 00:04:31.740 in any system. And it's all cash. And it's all 00:04:31.740 --> 00:04:34.899 cash. And now people say, well, we missed that 00:04:34.899 --> 00:04:38.680 2000 rand in our economic statistics. That is 00:04:38.680 --> 00:04:42.240 not quite true. First of all. Our street vending 00:04:42.240 --> 00:04:45.139 lady is buying her product somewhere. She may 00:04:45.139 --> 00:04:47.540 buy it from the fresh produce market. She may 00:04:47.540 --> 00:04:50.540 buy it from a fruit wholesaler. She may buy it 00:04:50.540 --> 00:04:52.839 from a next door neighbor who happens to grow 00:04:52.839 --> 00:04:55.500 tomatoes. That's quite possible. The next door 00:04:55.500 --> 00:04:58.160 neighbor is not giving it to her for free. So 00:04:58.160 --> 00:05:01.300 let's assume of the 2 ,000 rand sales that she 00:05:01.300 --> 00:05:04.720 had, that she had 1 ,000 rand in cost to procure 00:05:04.720 --> 00:05:07.730 the goods that she's selling. Then she has got 00:05:07.730 --> 00:05:10.370 some other costs. Let's assume that she's spending 00:05:10.370 --> 00:05:13.269 50 Rand on getting to the place, the spot where 00:05:13.269 --> 00:05:16.189 she's selling it. And let's assume that she's 00:05:16.189 --> 00:05:18.850 paying another 50 Rand to a youngster who helps 00:05:18.850 --> 00:05:21.069 her to carry the goods and look after the store 00:05:21.069 --> 00:05:25.889 for the day. So she earns 2 ,000 Rand. The basic 00:05:25.889 --> 00:05:29.189 input cost is 1 ,000. There's 50 Rand transport 00:05:29.189 --> 00:05:31.970 cost. There's 50 Rand to an assistant. That's 00:05:31.970 --> 00:05:37.100 100 Rand. She nets 900 Rand. And what you capture 00:05:37.100 --> 00:05:41.360 in the production accounts is the R1100. How 00:05:41.360 --> 00:05:43.899 do you mean what you capture in the production 00:05:43.899 --> 00:05:47.500 accounts? Well, the sales that are made by the 00:05:47.500 --> 00:05:51.980 fresh produce market or the sales that are made 00:05:51.980 --> 00:05:56.000 by the shop that sells the tomatoes. Her purchase 00:05:56.000 --> 00:05:58.899 is their sale. And so it will be recorded there. 00:05:58.899 --> 00:06:00.839 So it's counted on their books? It's counted 00:06:00.839 --> 00:06:03.680 on their books. Now, what we haven't explained 00:06:03.680 --> 00:06:06.699 yet is the 900 Rand out of the 2 ,000. What happens 00:06:06.699 --> 00:06:10.699 there? Well, that 900 Rand gets spent. Even if 00:06:10.699 --> 00:06:12.500 she puts it under the mattress for a week or 00:06:12.500 --> 00:06:15.319 two, at some point it gets spent, spent on her 00:06:15.319 --> 00:06:19.240 personal... food and clothing and transport the 00:06:19.240 --> 00:06:21.399 things that people spend money on that is after 00:06:21.399 --> 00:06:23.860 all why she's doing the street vending is to 00:06:23.860 --> 00:06:26.519 generate money for her own existence and the 00:06:26.519 --> 00:06:29.360 900 rand goes into the economy and it's picked 00:06:29.360 --> 00:06:32.339 up by the various people who sell her the services 00:06:32.339 --> 00:06:35.920 so if you if you look at it from these two sides 00:06:35.920 --> 00:06:39.300 you can see that most of the 2000 rand gets captured 00:06:39.300 --> 00:06:42.980 wherever it is getting captured but of course 00:06:42.980 --> 00:06:46.660 it's realistic to say that not every rand in 00:06:46.660 --> 00:06:50.100 the economy gets recorded. It is only realistic 00:06:50.100 --> 00:06:53.160 to say that some money will flow around that 00:06:53.160 --> 00:06:56.480 we do not capture. Now, how much? How much? That's 00:06:56.480 --> 00:06:58.839 the big question. I found it interesting that 00:06:58.839 --> 00:07:03.100 StatCSA themselves acknowledges in the way they 00:07:03.100 --> 00:07:06.259 present these figures that there will be some 00:07:06.259 --> 00:07:08.300 money which is not counted anywhere. Absolutely. 00:07:08.360 --> 00:07:11.620 And what's their thumbsuck? So their thumbsuck 00:07:11.620 --> 00:07:15.519 is 8 % of the economy is informal. And there 00:07:15.519 --> 00:07:18.100 are various ways, again, international benchmarks, 00:07:18.259 --> 00:07:20.620 what other countries are doing. Now, Joe De Beere 00:07:20.620 --> 00:07:24.199 made it quite clear to me that he thinks that 00:07:24.199 --> 00:07:27.139 may be a little bit too low. That could go up 00:07:27.139 --> 00:07:31.199 to 9 % or even 10 % of GDP. Now, when the next 00:07:31.199 --> 00:07:33.540 rebalancing of the economy is done, which is 00:07:33.540 --> 00:07:37.040 done every five years, they will probably up 00:07:37.040 --> 00:07:40.079 that 8%. But let's just work with that 8 % that 00:07:40.079 --> 00:07:42.579 we've got. Now, the interesting thing is if you 00:07:42.579 --> 00:07:46.300 look at 2023 numbers in the economy, that 8 % 00:07:46.300 --> 00:07:52.379 would amount to 562 billion rand. That is more 00:07:52.379 --> 00:07:55.300 than the entire mining sector in 2023, which 00:07:55.300 --> 00:07:58.399 only came to 444 billion. So just think about 00:07:58.399 --> 00:08:02.680 that. The imputed informal sector, 562 billion. 00:08:02.920 --> 00:08:05.939 The entire mining sector, that's platinum, gold, 00:08:06.199 --> 00:08:09.540 diamonds, coal, even stone quarries, all of that. 00:08:10.199 --> 00:08:14.600 only came to 444 billion. So the informal sector 00:08:14.600 --> 00:08:17.879 measured by this standard is bigger than the 00:08:17.879 --> 00:08:20.879 mining sector. And I think that should give comfort 00:08:20.879 --> 00:08:24.060 to people that we actually capture most of the 00:08:24.060 --> 00:08:27.779 data somewhere in the economy. You can fit a 00:08:27.779 --> 00:08:31.920 hell of a lot of informal transactions into 562 00:08:31.920 --> 00:08:35.419 billion rand. whether it's the tennis coach taking 00:08:35.419 --> 00:08:37.799 cash only, whether it's your doctor or dentist 00:08:37.799 --> 00:08:41.200 taking cash only, whether it's our street vending 00:08:41.200 --> 00:08:43.740 lady, whether it's a spa or shop, whether it's 00:08:43.740 --> 00:08:46.720 a brothel. The money goes out and somewhere it 00:08:46.720 --> 00:08:49.659 gets captured. So I think we can rest assured 00:08:49.659 --> 00:08:53.480 that by and large, we know we have a grip on 00:08:53.480 --> 00:08:56.240 how big the sector is. What does this mean in 00:08:56.240 --> 00:08:59.659 terms of unemployment? That's where things become 00:08:59.659 --> 00:09:03.200 very interesting. The unemployment numbers that 00:09:03.200 --> 00:09:06.639 got the Capitec boss animated is the fact that 00:09:06.639 --> 00:09:09.220 StatsSA said in the first quarter of this year, 00:09:09.340 --> 00:09:14.320 our unemployment rate is 33%, 32 .9%, make it 00:09:14.320 --> 00:09:19.080 33%, or 8 .2 million people. And against that, 00:09:19.179 --> 00:09:22.039 there were 16 .7 million people, more or less 00:09:22.039 --> 00:09:24.559 double the number, who were actually working 00:09:24.559 --> 00:09:28.440 and were employed. And he felt that these numbers 00:09:28.440 --> 00:09:31.080 are not right. Now, on what basis did he say 00:09:31.080 --> 00:09:35.360 that? He said that because Capitec has got 24 00:09:35.360 --> 00:09:40.080 million individual clients. They have 8 .8 million 00:09:40.080 --> 00:09:42.320 fully banked clients. I'm not sure precisely 00:09:42.320 --> 00:09:45.779 what that means. And they've got 13 million active 00:09:45.779 --> 00:09:49.419 users of the Capitec app. And he says those numbers 00:09:49.419 --> 00:09:52.779 tell him that the unemployment rate must be much 00:09:52.779 --> 00:09:56.490 lower. The problem with that approach is those 00:09:56.490 --> 00:09:59.830 numbers from Capitec do not tell us who is employed, 00:10:00.190 --> 00:10:03.129 who is unemployed, who is on a social grant, 00:10:03.409 --> 00:10:06.490 who is receiving a COVID grant, who is a pensioner 00:10:06.490 --> 00:10:09.289 that receives a pension but has got a Capitec 00:10:09.289 --> 00:10:12.230 account. You cannot draw just from that data 00:10:12.230 --> 00:10:15.450 conclusions about unemployment. I think that 00:10:15.450 --> 00:10:18.690 would be very wrong. So how does Stats SA do 00:10:18.690 --> 00:10:22.690 the unemployment measurement? They have 33 ,000 00:10:22.690 --> 00:10:26.169 households. which they survey every quarter, 00:10:26.370 --> 00:10:30.210 four times a year. And there you ask people questions 00:10:30.210 --> 00:10:33.490 around, have they earned money in the past week 00:10:33.490 --> 00:10:36.769 from working? Are they working somewhere? Is 00:10:36.769 --> 00:10:39.350 it a fixed full -time job or a part -time job? 00:10:39.490 --> 00:10:42.610 That sort of thing. And from these household 00:10:42.610 --> 00:10:45.730 survey, they then calculate the unemployment 00:10:45.730 --> 00:10:49.570 number. Now, the criticism is, first of all, 00:10:49.570 --> 00:10:53.049 that the sample of 33 ,000 houses It's based 00:10:53.049 --> 00:10:57.289 on the 2011 census, which is old, it's long ago. 00:10:57.750 --> 00:11:00.629 That's not quite correct. That number gets updated 00:11:00.629 --> 00:11:05.330 with regular other stats essay surveys. So it's 00:11:05.330 --> 00:11:09.169 not a number just from 2011. But is 33 ,000 enough? 00:11:10.029 --> 00:11:13.210 Yes, it is. You know, again, these things are 00:11:13.210 --> 00:11:15.210 done in terms of international comparisons and 00:11:15.210 --> 00:11:18.289 international benchmarks. And we're a country 00:11:18.289 --> 00:11:22.110 of about 60 million people. We do 33 ,000 households. 00:11:22.690 --> 00:11:25.330 It's comparable to a country like the UK with 00:11:25.330 --> 00:11:29.250 also about, what, 60, 65 million people using 00:11:29.250 --> 00:11:32.009 even fewer households. So, yes, it is enough. 00:11:32.250 --> 00:11:35.289 And what about the question, Mark, about the 00:11:35.289 --> 00:11:38.730 honesty of the respondents that people don't 00:11:38.730 --> 00:11:42.000 want to say? that they have earned money. Well, 00:11:42.080 --> 00:11:44.519 that's a good point. And that goes to the quality 00:11:44.519 --> 00:11:47.580 of the interview between the respondent and the 00:11:47.580 --> 00:11:50.899 stats essay official walking around asking the 00:11:50.899 --> 00:11:53.500 questions. But I think, you know, you must give, 00:11:53.639 --> 00:11:55.960 it's difficult to determine that, but I think 00:11:55.960 --> 00:11:58.460 you must give some credit to people who's got 00:11:58.460 --> 00:12:01.679 experience of this working around that problem. 00:12:01.759 --> 00:12:03.919 But it is quite possible that there's room for 00:12:03.919 --> 00:12:06.919 error. I think that is possible. So what do we 00:12:06.919 --> 00:12:08.960 learn from this? I think what we learn is that 00:12:08.960 --> 00:12:11.600 Capitec obviously sits on very valuable information 00:12:11.600 --> 00:12:15.360 and so do all banks. It's not just Capitec. And 00:12:15.360 --> 00:12:17.879 if one can combine the information that they 00:12:17.879 --> 00:12:21.460 have with StatsSA data, we may get a much better 00:12:21.460 --> 00:12:24.759 reading of the unemployment problem. StatsSA 00:12:24.759 --> 00:12:28.620 uses a survey method. Maybe one can supplement 00:12:28.620 --> 00:12:32.120 the survey method with data from the banks. Now, 00:12:32.179 --> 00:12:34.019 of course, you'll have to get the banks to cooperate. 00:12:34.259 --> 00:12:36.519 They'll have to make their data available on 00:12:36.519 --> 00:12:39.259 an anonymous basis and so on. It's not an easy 00:12:39.259 --> 00:12:42.379 thing to do, but there are people who feel that 00:12:42.379 --> 00:12:45.940 you can do that. And if the Gadifuri debate, 00:12:46.179 --> 00:12:49.000 if I can call it that, if it results in that 00:12:49.000 --> 00:12:52.379 kind of cooperation between the banks and Stats 00:12:52.379 --> 00:12:54.919 SA, I think it would have been worth all the 00:12:54.919 --> 00:12:59.440 carbon dioxide that was ejected. So in summary, 00:12:59.559 --> 00:13:01.980 so what? I think in summary, the first thing 00:13:01.980 --> 00:13:04.120 to remember is that you will never have 100 % 00:13:04.120 --> 00:13:07.480 accurate numbers. Forget that. The question is, 00:13:07.559 --> 00:13:10.980 can you be 90 % or 85 % correct? That is, I think, 00:13:11.000 --> 00:13:14.200 the benchmark to apply. The second thing is certainly 00:13:14.200 --> 00:13:17.419 that anecdotal evidence is not good enough. When 00:13:17.419 --> 00:13:19.659 you talk about something as complex as an economy 00:13:19.659 --> 00:13:24.799 involving 62, 63 million people to say, but I 00:13:24.799 --> 00:13:27.659 had this or that experience is not quite good 00:13:27.659 --> 00:13:31.480 enough. You're one of 63 million. And I think 00:13:31.480 --> 00:13:34.519 one should sort of display a bit of modesty, 00:13:34.580 --> 00:13:38.200 shall I put it that way. Thirdly, the system 00:13:38.200 --> 00:13:40.960 of national accounts which we run in South Africa 00:13:40.960 --> 00:13:43.980 is a robust system. It is applied internationally. 00:13:44.679 --> 00:13:50.220 It's applied in almost 200 countries. with similarities, 00:13:50.220 --> 00:13:53.059 but also differences with South Africa. So I'd 00:13:53.059 --> 00:13:55.480 be very careful to just chuck the baby out with 00:13:55.480 --> 00:13:58.379 the barf water. I think we can rely on what comes 00:13:58.379 --> 00:14:02.820 out. The current inclusion of the informal sector 00:14:02.820 --> 00:14:07.200 in our national accounts is 8%, and that is much 00:14:07.200 --> 00:14:09.700 bigger than the mining industry. And it's something 00:14:09.700 --> 00:14:11.960 just to think about. South Africa is a mining 00:14:11.960 --> 00:14:14.100 country, and there are a hell of a lot of mining 00:14:14.100 --> 00:14:17.000 activity. If the informal sector is bigger than 00:14:17.000 --> 00:14:19.740 that, then it tells you that we're probably not 00:14:19.740 --> 00:14:23.179 missing too much. We will miss some, obviously, 00:14:23.360 --> 00:14:26.799 but not too much. And lastly, I think there is 00:14:26.799 --> 00:14:30.820 room for cooperation between the banks who have 00:14:30.820 --> 00:14:35.139 the data of their clients available and stats 00:14:35.139 --> 00:14:37.720 essay and putting it all in a pot and trying 00:14:37.720 --> 00:14:40.100 to come to a better understanding. For example, 00:14:40.100 --> 00:14:42.840 if you take the banking data and you clean it 00:14:42.840 --> 00:14:45.220 up from people who receive grants, which you 00:14:45.220 --> 00:14:48.480 can do from ID numbers, it'll give you already 00:14:48.480 --> 00:14:51.080 a different picture. If you can clean it up further 00:14:51.080 --> 00:14:53.919 from regular payments, which reflect something 00:14:53.919 --> 00:14:55.960 like a pension, gives you a further cleanup, 00:14:56.120 --> 00:14:59.039 then I think one can come at useful figures. 00:14:59.620 --> 00:15:03.159 Thank you very much. That certainly taught me 00:15:03.159 --> 00:15:05.759 some things. I'm glad. That's good. Thank you. 00:15:07.659 --> 00:15:10.019 Thank you for listening to the So What Podcast. 00:15:10.399 --> 00:15:13.220 If you enjoy this content, please don't forget 00:15:13.220 --> 00:15:16.179 to leave a review and a rating, and please consider 00:15:16.179 --> 00:15:19.039 subscribing so you don't miss any future episodes. 00:15:19.320 --> 00:15:22.700 Also, tell your friends. Remember, you can find 00:15:22.700 --> 00:15:26.279 a written version of all JP's content at jplandman 00:15:26.279 --> 00:15:27.580 .co .za.