Should South Africa tax the rich more?

It is an interesting question. One that is no doubt supported by many people. In my view there is some scope in raising taxes for top income earners, but I think the same also holds for middle to lower income earners. If the objective is raise taxes to finance the deficit gap then raising the top income tax rate might not do much. Let us look at some of the data to understand why.

For the analysis I use SARS revenue collection data obtained from the National Treasury website. They break taxable income and assessed taxes by 24 income categories. The data is slightly dated - the most updated is in 2012. Some basic summary statistics of the data show that:

• Tax as a percent of taxable income is about 20.2% - implying that this is the average tax rate
• There were 14 million (m) taxpayers
• Only 5.8m were liable to pay income tax
• There are about 52m South Africans (See StatsSA 2011 census)
• Of the 52m, roughly 34m are of working age (although it is from ages 15-65)
• This means that about 17% of the working age population contribute to income tax

Given the high unemployment rate it is no surprise that only a small percentage of people contribute to income tax. We can calculate the effective tax rate of the income groups (this is plotted below). It is clear that there are two tax slopes: R60,000 - R150,000 and one from R150,001-R5 million and above.  This does not correspond to the hump-shaped statutory tax rate schedule. This is interesting for two reasons: i.) that some people earn additional income other than salaries and wages such as dividends and interest that cements these linear slopes, or ii.) it could mean that there is the possibility to introduce additional tax thresholds (this often over-complicates the tax system).

We want to compare these rates to the proportion of total revenue collected by income group and analyse how this stacks up to the proportion of tax payers. We see that the majority of income is collected from people who earn R150,000 and above. This corresponds to the number of tax payers in that income category (the number of tax payers above R1.2 million decreases significantly). About 86% of total revenue collected comes from the wealthy 30% of people - meaning that 70% of people only contribute about 14% of income tax revenue collected.

Just to see how unequal the tax distribution is we calculate the Gini coefficient. A measure of 1 means absolute inequality and a measure of 0 implies absolute equality. We compare taxable income with tax assessed (see Figure below). There is a more unequal distribution of collecting taxes compared to taxable income with a Gini of 0.652 vs. 0.558. A heavier weight from on the 50th to 70th percentile vs. what they actually earn. This does not imply that these people pay more than what they should (how much each should pay is a philosophical and economic question left for a later blog entry). It means that there is a decent sized middle-class that pays taxes.

But just how much additional tax can the government collect from simply raising tax rates? It is very difficult to get a decent estimate. If there are no adverse behavioural responses to an increase in tax rates then the calculation is simply mechanical, i.e. take the lump sum amount add the amount of revenue above the threshold times the rate. Unfortunately there are many behavioural effects of taxes. People may not work as hard as they used to (despite the same working hours), lower income households might stop working altogether if they are able to receive better benefits from being unemployed. 

In the next post we will look at some of these effects and try to arrive at preliminary conclusions.

For now it suffices to know that the tax burden is disproportionately higher than on a small number of individuals. We know that tax inequality is slightly higher than income inequality (the middle class pay a slightly higher share of taxes compared to their incomes relative to the rest of the income distribution from the SARS sample data). 

What is that - we are giving more money to Eskom?

The South African government's policy on Eskom is interesting. Definitely not in a good way.

Eskom has received about R60bn in loans from government not too long ago. Now government is giving Eskom another rescue package. And, what has Eskom achieved? It seems like very little but empty promises. The delivery of power stations have been delayed. New project costs are well above estimates. Makes me wonder what project finance plans these guys have in place and why no one is being held accountable for crippling the South African economy. It is definitely not fine that more people are out of jobs because of bad economic planning. It is a crime, and it seems like a perfect one since no one is held responsible.

There are two very funny parts about the Eskom saga: People pay taxes to bail out Eskom. They don't get free electricity for this. On top of that, they have to pay Eskom's tariffs. And then on top of that, they have to pay municipal taxes that add a markup on Eskom's tariffs. These markups differ from municipality to municipality. It seemed like the Treasury was swayed by a possible downgrade. The ironic part about a bailout is that it will put more pressure on the government's finances and could lead to a possible sovereign downgrade.

What does all of this mean? While I don't think that South Africa's debt is such a big problem in comparison with many other countries, it seems like rating agencies disagree (they have been wrong in the past...hint: financial crisis and mortgage backed bonds). It means that there will be an increase in government's effort to consolidate its deficit and reduce its debt. If not, brace yourselves for a downgrade. These are the consolidation options government has:

1. Decreasing government spending. What will they decrease? It will be stupid to cut back on infrastructure spending since this is the only component of SA's government spending that is productive (i.e. there is some future benefit that can be discounted). They will struggle to decrease the wage bill - employment is largely supported by government. They will struggle to cut real wages since they will face formidable opposition from unions. Health? No. Education? No. So it seems like government will not really cut back on spending.
2. Raising taxes. I do not think it is the right environment to raise corporate income tax. The economy is still too weak, and businesses will not stomach this (and yet we have corporate profits doing just fine...?). There are also some papers that argue that an increase in corporate income tax only temporarily decreases growth. But we can't really afford a reduction in GDP at the moment. Perhaps there is room to raise taxes for the upper income households. Of course they won't like this since they already contribute the majority of SA's PIT collection. And the additional revenue collection from raising the upper threshold tax rate will be too little. Perhaps government will finance the Eskom bill and the wonderful wage bill by increasing VAT? To be honest, I think this is the most real response to their problem. It is, however, political suicide. But that should not dissuade the ANC too much now - they have recently won the general election.
3. A miraculous increase in economic activity that does not rely on electricity. Electricity constraints has played it share in a slumpy growth environment. So where will this miraculous growth come from? 
4. Inflation. Nope...SA will not be able to inflate away its debt problems. The inflation increases will definitely be outweighed by interest rate increases. South Africa is an inflation targeting country after all.

Unfortunately, it seems that the government is left with no choice but to bail out Eskom. But this time at least fire people who are not doing their jobs. Make sure Eskom employees do not get crazy salaries. And dammit, make sure that pay raises are linked to productivity growth and not some stinking measure! Make sure there are proper loan conditions. And open the debate for competition. The days are gone when South African electricity is cheap. A single supplier of electricity has market power - and market power in the hands of evil men mean high prices. And yes they can charge super high prices since most of us need electricity (the price elasticity of demand is very inelastic).

South Africa's unemployment curse

We know that SA unemployment remains stubbornly high. Despite various policy changes over the last twenty years not enough has changed. In this post I point the blame-finger at the lack of good education.

The current unemployment problem in South Africa reminds me of England during the 1800's: Sylvia Nasar paints a rather vivid image of England in her book Grand Pursuit - a rich nation stricken with poverty. Unemployment was high, people were uneducated, communities were poorly developed and it seemed like the rich were simply getting richer at the expense of cheap labour. This last point does not necessary hold for South Africa. While the British are going through some economic turmoil currently, the average life of a single human being is much better than a South African. Somehow the English have escaped the 1800's problems, and somehow current-day South Africa is facing a situation that could resemble an 1800's England 

As I see it there are mainly three problems that broadly explain unemployment in South Africa (there area lot of good academic papers on this subject - I don't intend to contradict or necessarily improve on that work):

1. Labour is unskilled
2. There is a wage premium mismatch
3. The structure of education is brittle and outdated

I will discuss my views on each point separately (they are definitely related too)

(1) Labour is unskilled

Labour is unskilled because the quality of education is shocking. South Africa, as an example, scores very low in mathematics and science relative to other countries (just look at the Pirls and Timss results). Why? We don't have enough teachers, we do not have skilled teachers, we do not provide proper study materials for students, infrastructure is nonexistent in many rural areas (although I think that this is the least of the problems, still severe though), only the rich can afford good private schools etc...

Why don't we have enough teachers when South Africa's current greatest economic problem requires us to have many more skilled teachers? This is a simple economics problem. If there is such a high demand for teachers, then why don't wages increase in proportion to the demand (assuming there is a shortage of the supply of teachers)? Or is real growth in teacher wage bill capped by the quality of teaching they offer?

One would think that SA's high education expenditure would solve these issues, yet there is little change in educational outcomes. One of the things I am curious about is how much of the education spending actually goes into teaching versus the administrative aspect. I.e. by how much do wages change for government employees administrating education vs. wages for actual teachers? I am also curious about a comparison of the employment uptake for teachers vs. administrators in government. 

Expect a bleak South African future if the current problems in education are not solved. A low knowledge base exacerbates the problems of the unskilled unemployed and reduces economic growth. Slower future growth will result in more unemployment and most probably increased political turmoil.

(2) - Wage premium mismatch

I often ask myself why critical skills (engineering, education, medicine, science) wages are on average way below wages of finance and management executives. The word critical for me would imply a shortage of these skills or super high demand for these skills - as a result real wages should rise by a lot. The problem is not about critical skills alone, we do not seem to have the ability to absorb such labour and create industries that utilise such labour effectively? South Africa seems to outsource a lot of critical infrastructure projects to foreign nations while perfect solutions could exist back at home. As a consequence we do not utilise the critical labour as much as we should and hence this could explain in part why real wages in these industries are not comparable to finance.

The wage premium mismatch is also exacerbated by South African unions. Let me first state that I think unions serve a valuable economic function and are necessary. My problem with unions in South Africa is that they take an explicit communist stance without any real cause. Marx prophesied that capitalism will cause a drop in real wages over time and result in longer working hours where the lives of people become more miserable. Real wages across sectors have been rising since the fall of apartheid and working conditions have improved for labour. It is true that income inequality has risen. My view is that income inequality is more important for social stability than anything else. Household income and poverty are a more important variables to look at when comparing how people's lives have improved over time - poverty and household income across all income deciles have increased. I feel rather bemused at the crazy wage demands that unions make given that real wages have risen over time, poverty has decreased and household income has increased. Another concerning factor is that the demands for an increase in real wage growth for unskilled labour seems to outpace wage growth for skilled labour. I am not saying that this is wrong - since unskilled labour work just as hard as skilled labour and often suffer physical consequences for that labour; but it does violate certain economic principles and has the potential to destabilise the economy further - I don't think anyone would dispute that union unrest over the last couple of years have hurt the SA economy.

This brings us to out final problem - and a proposed solution

(3) - The structure of education is brittle and outdated

We know that the current curricula is something to laugh at and does not provide our children with the best opportunities for life. We also know that circumstance plays an unfair role in deciding what type of education a particular child receives - wealthy parents are most likely to send their children to top private schools while poor parents have no choice but to send their children to a rural school where teachers do not even show up for work. Apart from these choices, malnutrition and creating a safe environment are also important factors that help explain the success of education.

If we know that this is a problem, perhaps the greatest challenge that a democratic South Africa is faced with, why isn't the government and private sector doing everything to change this? The effort just seems so feeble and pathetic. The myopic nature of government and the private sector will create social and economic problems for ages to come.

Some solutions:

First, education should be provided for free to give children equal opportunities to go to decent schools. Second, university graduates would do well if they spend 6 months to a year being placed in a school to teach an accepted good quality curricula - some people might be unhappy with this, but it is a low cost effective solution and should not be too difficult to implement. Third, teachers need to get a good education and incentives should be put in place to become teachers or lecturers that train teachers. This could be in the form of higher wages, better subsidies etc.

Once this is in place one can start to revamp the school system in its entirety. All students receive the same curricula up until grade 7. From grade seven children get placed. Children who score above a certain mark (let us call this X1) are placed in classes that do proper training in mathematics, science, history and language - these children will be eligible to go to university and study courses that fill critical skills. The children's incentive is that they get to choose their study path. Children who score below the above threshold, but obtain still reasonable marks (X2), i.e X2<score<X1, are also trained in the sciences and are also allowed to go to varsity - but they are not allowed to study engineering, science or medicine. They still have an incentive to work hard - they get to go to university. Finally, children who score below an accepted threshold (X3) are placed into various training colleges that teaches them skills as artisans, administrators, construction and other services. This way we make sure children obtain skills (something that is presently lacking) and are placed in jobs where there is a demand. It will also encourage children to work harder if the do not want to work in these industries.

I really do not see an alternative to the pressing education gap in South Africa. The entire school system is not working and needs to be completely overhauled.

Obviously this is my take on part of the unemployment-skills problem in South Africa. I have no doubt that there are many holes in my arguments. I urge you to help come up with a solution that is better than this, something that is realistic. Or, if you spot serious gaps in my approach, fill them with details that improve it. If we all want to live happily ever after in South Africa then we all need to take this problem very seriously.

In the next couple of blog posts I will try and provide some data on some of the problems I highlighted. I would like to compare real wages across jobs (unfortunately the labour categories in the labour force survey are rather limited). If possible, I would like to compare the wage bill for teachers vs. the wage bill for education administrators in government. I would also like to know how much of of our infrastructure projects are outsourced to foreign national companies vs. domestic companies - can we explain the differences? Finally it would be interesting to get a view on how many critical skills positions are advertised as opposed to not so critical skills - hopefully this will give us a view on the demand for these skills and possibly analyse whether there isn't any mismatching problem.

For some bedtime reading on the huge returns of education on economic growth and reducing income inequality read this and this.