My previous article showed that South Africa has more employees than it used to. Statistics SA revises this data every year, and has found some 16.4% more people since 2014.
That is a lot of extra employees, and they earned more income too.
Yip, they earned an average of R113.4 billion more every quarter last year, or about R453.6 billion more for the year. Quarterly employment statistics (QES) income for formal sector employees is 19.2% more than the GDP data shows.
That alone is nearly 10% of GDP (as employee compensation makes up about half of GDP), and that is the difference in the formal sector alone. If the informal sector were added to the QES income numbers, then the difference to be added would be even bigger.
The QES does not measure the informal sector. The informal economy is at least 7% of SA GDP. So add 7% to the 19.2%, and you end up with 26.2% more value added via employee income. Some estimates put the informal sector at 13% or so, which would imply about a third more employee income.
The above is not a shock for many as the data comes from official sources – in fact, the data comes from exactly the same source that publishes our GDP! But most of this data is based on administrative sources such as company and government payrolls and backed up by Unemployment Insurance Fund (UIF) data.
Early UIF data also supports the QES data and is essential as it supports other indications that the SA economy is larger than GDP data suggests.
The graph above illustrates how much more the total earnings figure in the QES rises as a percentage of the earning recorded in the GDP. (GDP earnings are for all non-farm employment and QES data relates only to the non-farm formal sector, so the difference will clearly be far higher.)
This data is also backed up by large increases in tax collections, particularly value-added tax (Vat) – the rate of which remained constant for a number of years during the 1990s and early 200os.
A larger workforce and compensation is backed up by the massive increase in the number of vehicles in South Africa. Cars and bakkies combined are basically equal in number to the formal non-farm workforce.
Furthermore, the number of malls that SA sustains as measured in square metres is one of the highest in the world. Not quite as would be expected from a developing country (SA has the second most square metres of mall space per person employed which is incredible for an emerging market).
Indications, therefore, are that SA GDP is underestimated by at least 10%. Perhaps even closer to 20%.
If GDP is so much larger, why did we not pick this up earlier?
A while ago an international tax expert told me that it was beyond amazing that SA tax collections are estimated to have lower leakage than most advanced countries. He told me to look at the increases of tax to GDP, while every year the South African Revenue Service (Sars) tells people it misses less than 4% of tax every year.
Yet, as he pointed out, year after year for about two decades tax collection came in above estimate and Sars again states that it catches nearly everyone – yet more firms and people are again caught in the tax net the next year.
You are an emerging market, he said. You should have a bigger informal sector with less efficient collection of taxes – that is why emerging markets collect more indirect taxes. Your focus should be there. SA keeps on overachieving on the tax collections. That is not ‘normal’ for a developing country.
This conversation just sat in my mind for years.
How did we achieve these continuous record-breaking tax collections and have tax-to-GDP roar up while personal income tax rates declined for a decade?
Should we have woken up and smelt the roses earlier? For sure!
Yes, some of the increases in the tax-to-GDP ratio were due to better collections but I now know that at least some of that extra tax collected was due to better economic growth.
The Vat rate did not change for decades, and collections as a percentage of GDP increased in the period from the end of the 1990s to 2008/9. The increases in Vat-to-GDP are shown below. In 1995, Vat collection was under 5% of GDP. It peaked at over 7% and ended around 7% although the rate of Vat never changed.
While SA bragged about how efficient Sars was, lately we have noticed how quickly tax slippage can happen and may now understand that it was GDP growth that also helped tax collections.
Like Eskom, Sars probably had a little skin in the game if it did better than expected with collections. That suggests that economic actors were never as compliant as we were led to believe, but they classified it that way.
So, the fact that SA probably never quite plugged all the tax leakage is the real answer to the tax experts’ question.
Why is the larger number picked up now?
As employees become more knowledgeable about UIF and their rights to it, they reported non-compliance far better than Sars. They ask ‘Where is my UIF?’ at labour offices all over the country, and non-compliance is exposed. Employers had to become compliant, or the smelly stuff was going to hit the fan.
Employers may still under-report some earnings, but with UIF they no longer have a chance to do so; half a million people applying for UIF every year makes non-compliance far harder.
Effectively, the retrenched found the ‘lost’ economy.
That international tax expert, considering the advice he gave, knew that Sars was not as good as it was portrayed.
He knew something did not make sense when we claimed less non-compliance than developed countries with massive administrative data sets that make tracing it easier.
Since when is the GDP level higher?
I believe the GDP growth trend has been correctly assessed, but the actual level of GDP has probably been far higher – even as far back as the early 1990s.
SA’s big growth was taking place during the commodity boom, and it was our primary sectors, with their ability to export, that was the backbone of that period of growth (and remains so).
There is just no way that the formal sector in South Africa would have created over 1.4 million jobs since 2014 as business confidence was negative and our GDP growth was low – in fact, our GDP per capita declined.
Since September 2014 real GDP has grown 6.6% while formal sector employment has risen by 16.4%. Normally, GDP growth would be far higher than employment growth. Over decades and across many countries there seems to be about 2% or 3% GDP growth for every 1% employment increase.
Stats SA got the general trend of GDP right, but not so the real level or actual growth rates.
The chart below shows the difference between the revised QES earnings and GDP compensation when expressed as a percentage of GDP.
In simple terms, it is impossible for both to be right, and I believe it is the larger one that more correctly portrays formal sector earnings.
If that is the case, then South Africa’s GDP is at least R450 billion larger than we knew before (I will gladly accept the 10% finders fee …).
Mike Schüssler is chief mischief-maker at Economists.co.za, but he is very serious about the underestimation of SA GDP.