An analysis of the tax proposals in the 2018/19 National Budget by Imraan Valodia and David Francis (Daily Maverick), 27 February 2018, South Africa
The budget is about more than just income and expenditure. It should provide a vision for the economy, which deals with the current challenges but also offers an insight into the type of society we want to build. The key message in the 2018-19 budget is that South Africa’s public finances are in very bad shape, and we have to find more revenue. The budget proposes that we all need to tighten our belts and make sacrifices to get out of trouble and to get the economy on a sustainable growth path. It asks all citizens to pay more, and all income groups to share in the burden. At the same time, however, the Treasury has decided to leave corporate tax unchanged.
As South Africans, we should all be asking the Finance Minister: “Why do only some of us have to make sacrifices? And, why can the sacrifice not be distributed in an appropriately proportional manner and include company taxes?” We believe that Business, in the interests of a holistic growth and public finance strategy, should have proposed and supported an increase the company tax rate. Thankfully after many years of silence, leaders in business have been outspoken about the challenges in our society and have made commitments to work toward addressing the high levels of inequality in our society. But talk is cheap, and it is easy to sign a pledge when little is at stake. Unless Business gets serious about joining everyone else in making sacrifices, and taking on its appropriate proportion of the sacrifice, one can’t but help wonder whether it is all talk and no action.
The budget could have been a mechanism to build a consensus across class divides which continue to undermine the sustainability of the economy. The political moment was perfect. An increase in VAT and an additional R13 billion in revenue from a 2% increase in the corporate rate would have demonstrated that the private sector is committed to accelerating sustainable and equitable growth. It would have been a small sacrifice for business and would have gone a long way to generate more revenue, and so creating some scope to better address poverty on the expenditure side of the budget. As it turns out the 2017/18 budget will be remembered for the first VAT increase in post-Apartheid South Africa. It should have been the budget remembered for all social groups, including Business, coming together to address our economic challenges.
Generating revenue is a delicate balancing act, and raises important issues of equality and fairness. This is why it is critical that both policy decisions and public debate are informed by evidence. We use the available data to make three arguments. First, given the dire economic situation in South Africa, the decision to raise VAT deserves careful consideration. Second, we argue that any evaluation of the tax proposals has to look at all of the available tax measures, including personal income taxes (PIT) and company taxes. In our view, the increase in VAT should have been accompanied by an increase in the rate for company taxes, which is currently 28% and has been falling. Third, our view is that the message of the budget should have been that collective sacrifice from all taxpayers is required to stabilize public finances. The burden to do this must fall disproportionally on those that have a greater ability to pay – the rich and the private sector. We argue that the government has failed to take advantage of a political moment which we believe would have allowed it also to raise the corporate tax rate. But these decisions are not only political, and here we consider the economic evidence.
The South African government has three main tax policy instruments – PIT, VAT and company taxes, which respectively make up 37%, 25% and 18% of total tax revenues. In this article, we discuss two other taxes: excise taxes and fuel levies. Excise taxes are taxes on specific goods such as alcoholic beverages, soft drinks and cigarettes – the so-called sin taxes. The tax is included in the price of the good when it is purchased. Similarly, the fuel levy is a tax that is included in the price of petrol and other fuels. Excise taxes and fuel levies, respectively, contribute 5% and 3% of total revenue. The biggest contributors to tax revenue are PIT, VAT and company taxes and we should, in evaluating the tax proposals, focus on these. In the tax literature, we often refer to direct taxes and indirect taxes. Direct taxes are what we pay directly to SARS: PIT and company taxes. An indirect tax is a tax that is paid when something else is purchased and the tax is included in the price of the good purchased. The main indirect tax is VAT. Excise taxes and fuel levies are also indirect taxes.
One of the key considerations in evaluating taxes, is the distributional effect –, whether the burden falls on the rich, or the poor, or both. We concern ourselves both with how much each group pays, and how much of the burden falls on each group, relatively to their ability to pay. The tax is progressive if the burden falls mainly on the rich; regressive if the burden falls mainly on the poor; and neutral if the burden is spread equally across all classes. A tax should not make the level of inequality worse – so a regressive tax is a bad policy. Government should be using the budget to improve income distribution and skew policies to favour the poor.
Ideally, one should look at both taxes and expenditure at the same time to assess whether the budget is progressive, regressive or neutral in terms of inequality. To keep to the main issues that concern us here, we only discuss the revenue side of the budget. However, it is important to note that because of social grants and other social support in South Africa, the expenditure side of the budget is strongly progressive. And, even though we have problems with state capacity, our government successfully pays out the social grants. The principle is that if poor households are hurt by any taxes, they should be compensated on the expenditure side of the budget by above-inflation increases in social grants.
The key tax policy proposals in the 2017/18 budget are: an increase in VAT from 14% to 15%; slight amendments were made to PIT, to protect lower-income groups against inflation by decreasing the amount of PIT they pay; and, the company tax rate was left unchanged at 28%. Excise taxes and fuel levies were increased.
First, let’s assess whether, theoretically, VAT is regressive or progressive, and whether it does place an unfair burden on the poor. For tax evaluation purposes, we concern ourselves not only with how much different groups pay but also with how much, proportionately, different groups pay; that is, what is the tax burden relative to the ability to pay. Let us compare two citizens, A and B. A, a low-income South Africa earns R30,000 per annum and spends all her income because she cannot afford to save. B, a high-income South African, earns R3 million per annum and, being wealthy, spends 50% of her income, saving and investing the rest. For simplicity, we assume that all of A’s and B’s expenditure is on goods that attract VAT. At 15% VAT, A pays R4,500 in VAT (15% of her income) while B pays R225, 000 in VAT (7.5% of income). In this scenario, VAT is regressive because, although A contributes significantly more to the VAT pool, relative to her income, A is paying more than B. A is paying 15c in each rand earned while B is paying only 7.5 cents in each rand earned. In theory, and in this simple scenario, VAT has to be regressive because it is a tax on consumption and, compared to the rich, the poor consume larger proportions of their income. However, in reality the impact of VAT is determined by how it is implemented and what goods are included in the VAT net and which not. In South Africa, we have extensive zero-rating of basic goods, and this has a big impact on whether or not VAT is regressive.
Table 1 shows the annual income earned by households by income classes in 2011 prices. It is calculated by ranking all households in South Africa from the lowest income household to the highest and then dividing the households into ten groups with an equal number of households. This table demonstrates how unequal income distribution is in South Africa, where we have a few very rich people and many poor people. In the poorest decile, on average, households earn R 17,534 per annum; about R1,461 per month. In richest decile, on average, households earn R517,323 per annum, about R43,110 per month.
Let us then discuss what the different income classes share of spending, VAT and other indirect taxes looks like. The left-hand panel in the table below shows us almost all we need to know. The poorest South African have 0.54% of disposable income and their share of VAT is 0.47%. The richest South Africans have 56.66% of disposable income and pay 59.56% of VAT. The panel on the right shows cumulative shares. So, quintiles 1-5 have 7.32% of disposable income and their share of VAT is 6.55%. The increase in VAT, from 14% to 15%, is expected to raise R22.9 billion. In total, quintiles 1-9 will pay about R9.26 billion (43.34%) while quintile 10 will pay R13.63 billion. The poorest quintile in contrast will pay just over R1 million. Clearly, the rich will pay a lot more of the bill in absolute terms.
But as we said earlier, we shouldn’t only ask how much each of the deciles pay. We should in fact be more concerned about their tax burdens relative to their ability to pay: how much of VAT does each of the deciles pay relative to how much they earn?
The table below shows us how much South African households spend on indirect taxes as a proportion of their income – VAT, excise taxes, and fuel levies. For the purpose of assessing the impact of VAT, we must focus on the blue bit of the bar graph. The blue bars are increasing slightly but consistently from left to right, from 9.5% to about 12% - this means that the poorest pay 9 cents in each rand of disposable income they earn, while the richest pay about 12 cents in each rand of disposable income they earn. Each of the income classes pays a little more than the class that are poorer. The conclusion is that VAT, in South Africa, is not regressive. Rather, it is very mildly progressive. It is not a great story (we would like taxes to be strongly progressive) but empirically it is not the case that the increase from 14% to 15% will place a disproportionate burden on the poorest South Africans. To be clear, poor households are worse off from having to pay the increase in VAT but rich households are proportionately more worse off. But as long as the taxes collected are spent on poor households the poor will still be better overall and inequality should fall.
The popular story in the debate on VAT in South Africa is that it is regressive. As we have demonstrated above, VAT is in fact mildly progressive, but this is not true for all indirect taxes. In the graph above, the blue portion of the bar shows the incidence of VAT, the red the incidence of excise taxes, and the green the incidence of fuel levies. Taken together, VAT, excise taxes and fuel levies are regressive – the poorest pay 17c in each rand, while the richest pay about 16c in each rand of disposable income. This is regressive because the poor are paying more tax than the rich for each rand of disposable income. However, it is excise tax, not VAT that causes the regressivity. The poorest decile earns only 0.54% of income, but pays 3.44% of the excise tax. In comparison, the richest decile earns 56.66% of income and yet contributes only 29.21% of the excise tax. In other words, VAT and fuel levies are not regressive. Excise taxes are strongly regressive. We should note that this evidence is undisputed and all studies of the incidence of VAT in South Africa have reached this conclusion. VAT is not regressive in South Africa because the zero-rating of basic goods does a very good job to protect the poorest South African.
Excise taxes are regressive because relative to their disposable income, the poor spend disproportionately more on goods that attract excise taxes – cold drinks, cigarettes and alcohol; the so-called sin taxes. The budget proposals to increase these sin taxes will place a further and disproportionate burden on the poor. This is a complex issue: quite clearly, the commonly held view that increasing sin taxes is good policy raises complex distributional issues. The budget has also increased the fuel levy. Taken together, these proposals on indirect taxes do in fact place an undue burden on the poor but it is important to distinguish the effects of each of these taxes.
Having clarified that VAT in South Africa is not regressive, it is worth investigating why the Treasury may have chosen to increase the VAT rate over the other tax options available. Tax authorities often talk about the three E’s in tax policy – equity, ease of administration, and efficiency. We have dealt with the equity consideration (though there is a bit more on that below); VAT is not regressive. From an ease of administration perspective, VAT is a transparent tax (we all know how much VAT we pay). From an efficiency perspective, there are two considerations – how much money does the state collect for a small change in the rate of tax, and how much of what the state should collect does it actually collect (i.e. how easy is it to evade the tax). VAT has great advantages on both these scores. Studies that have shown that the gap between what should be collected and what is actually collected is very small indeed – its much smaller in South Africa than in comparable countries. For these reasons, VAT is an extremely attractive option when it is important to quickly and cost-effectively raise revenue. This is especially so if there is a well-functioning VAT system, with a zero-rating provision that protects the most vulnerable citizens from the pernicious characteristics of VAT.
In total, through the adjustments announced in the budget, government is raising an additional R36 billion in tax revenue. The bulk of this, R22 billion, is expected to come from the increase in the VAT rate. From the perspective of the 2017-18 budget, the problem with VAT in South Africa is not that it is regressive, but that it is not strongly progressive. In other words, because VAT is by far the largest proportion of the tax adjustment, the overall impact of the adjustment is spread more or less equally among all income classes.
In a country that has a relatively equal distribution of income, it may be acceptable for the burden of an adjustment to be borne equally among all classes. South Africa, however, has a distribution of income that is among the most unequal in the world. The top decile in South Africa earns 60% of all the income and owns 95% of all the assets. The objective of economic policy, if we are trying to build a sustainable and humane economy, should be to change this unequal distribution of income and assets. Conceptually, it is quite simple. Addressing inequality requires policies that will increase the level of income for low-income groups at a rate that is greater than that for middle and high-income groups. This requires that the costs of economic stabilisation and adjustment are borne disproportionately by higher income groups. Essentially, it is a question of ‘proportionality’ – in our context if any burden is to be fair, the rich should, proportionately, pay a lot more than the poor.
How could the tax proposals have been better crafted to be more equitable? There are two other sources of taxes – personal income tax (PIT) and company taxes. We briefly assess each of these in turn. In the 2017/18 budget government is making only minor changes to PIT by adjusting tax brackets, to protect tax payers against inflation for low income group only. By not adjusting the rates for high income earners, government will raise an additional R6.8 billion.
There are two problems with PIT in South Africa. First, low-income groups fall outside the income tax net, so anything that government ‘gives back’ to the payers through the tax system (for example, by reducing tax rates), excludes the poorest groups. The tax threshold is just over R75,000, so anyone who earns less is outside the tax net. Second, because of the highly unequal distribution of income relative to the size of the population, we have very few taxpayers in the highest income groups – there are only 109,783 individuals in the highest income bracket. Raising tax rates for high income earners affects a small group of tax payers and therefore does not generate a large amount of revenue. In deciding not to make any further changes to the PIT, government has argued that in the last budget government added a new top income tax bracket of 45 per cent for those earning above R1.5 million. Furthermore, the increase in VAT affects all individuals when they spend their incomes; on balance, there was probably not a lot that could have been done in the short term about PIT.
A number of commentators, especially leaders in business who have commented on the budget, say that while the VAT increase is undesirable, it is the best of the options available. They go on to say that it is a very good thing that the tax rate on companies, currently at 28%, was not increased. The arguments made to support this are varied, and deserve investigation. Firstly, we need economic growth, which requires higher levels of investment by businesses, and raising the company tax rate will not promote new investment. In the Budget Review, government makes the argument that most of our trading partners have been decreasing the rate of company taxes, which limits the space for South Africa to increase corporate tax. Both of these are fair arguments. But there are arguments on the other side too. A 2% increase in company taxes would have generated about R13 billion.
Economically and politically it may have been a good thing to raise company taxes to 30%. That South Africa’s economy is highly concentrated is not debatable. Companies that operate in highly concentrated monopolistic markets earn high ‘monopoly profits’. These profits should be taxed at higher rates. In addition, no other country has an income distribution like ours and not many countries have such high levels of economic concentration. This requires action beyond joining the “race to the bottom” in corporate tax rates. Furthermore, the investment argument above has some flaws – investment is driven not only by tax rates. The standard of infrastructure, policy certainty and the like are more important considerations. Unlike VAT, where the rate has remained unchanged in post-Apartheid South Africa, the company tax has been decreasing since the early 1990s, when it was at 50%. The argument that decreasing the company tax rate will boost private investment has been made since the early 1990s and throughout the period, private investment has been low and falling. In our view, the burden of the fiscal adjustment should be shared among all taxpayers including the corporate sector and there are just as good economic arguments for an increase in company taxes.
We believe that the Treasury has failed to capitalise on a political moment which would have allowed it to raise the corporate tax rate. An increase in the corporate tax rate would have positive implications for the distribution of the tax system, and would have ensured that those most able to contribute to plugging the hole in South Africa’s public finance are the ones that carry a higher burden of tax. We also argue strongly for a progressive tax architecture. But for this to be achieved, we have shown that VAT is not in fact regressive, and that the highly regressive nature of excise tax must be addressed. Finally, we must emphasise the in South Africa we are fortunate to have excellent data that make evidence-based discussions about these important issues possible.