Higher tax rates in South Africa

Most countries have increased their sovereign debt levels since the 2008/09 financial crisis to stimulate slowing economies or bail out the greedy bankers who tried to make more money for greedy investors. It is a sad thing really. Now governments are facing probabilities of default (just type in sovereign debt crisis EU in google to understand just how bad things are). To minimise the probability of default governments have to increase tax rates, cut rebates or reduce spending. All these measures have negative consequences for economic growth. The point is that some governments simply don't have a choice as capital to finance expenditure dries up. The Obama administration is letting the Bush tax cuts expire (which effectively means that tax rates increase http://www.thefiscaltimes.com/Columns/2012/12/04/Why-the-GOP-Wont-Admit-Supply-Side-Econ-Has-Failed.aspx#page1). Unfortunately the question about which tools to use to appropriately consolidate the budget deficit is a rather complex one. In economics we learn that fiscal spending and taxes have multiplier effects (i.e. a dollar increase in spending increases economic activity by x dollars). If taxes have a lower multiplier than spending, then increasing taxes will harm the economy less than cutting spending. But calculating the multiplier is a very difficult task and one cannot simply rely on any old estimate. However, there seems to be empirical support that taxes have smaller multipliers than spending.

Let us assume that taxes damage the economy less than spending cuts, what taxes do we change and what rates do we set? This question is also very complex. One has to consider the redistributional preferences of government, the public's aversion towards income inequality, the possibility of tax evasion and tax avoidance and the possibility of immigration. Most undergraduate textbooks show that lump-sum taxes are Pareto efficient (the point where an increase in welfare of one individual cannot take place without decreasing the welfare of another). This tax strategy also assumes that government is efficient in distributing revenue collection efficiently. Unfortunately inequality exists and governments are far from efficient. This is why we have progressive taxes which addresses vertical (people with higher incomes should pay more taxes) and horizontal (people with the same income should pay the same amount of taxes) equity concerns. This unfortunately implies a Pareto-inefficient outcome as the tax liability lies with the minority of the population. This is very true for South Africa!

So, is it at all possible to increase the tax rates of the tax paying population further? Can these people afford a higher tax rate? Most probably yes. It is important to know that there is a difference between the effective tax rate an individual pays and the statutory rate. If person A earns R600 200 per annum he would fall in the top tax bracket. Assume that the cut-off point is R600 000 with a lump sum tax of R100 000 and 40% tax of anything in excess of R600 000. His total tax liability would then be (R100 000+(R600 200-R600 000)*40%) R100 080. Assume that the tax rate would increase to 50%, then his liability would be R100 100 which is only a R20 increase. His effective tax rate before the change is ((R100 000+R80)/R620 000) 16.141%. His effective tax rate after the change is 16.145%. That is why you always need to look at the change in rebates, tax rates and lump-sum taxes when the Budget is revised.

We know from theoretical and empirical (see Mirrlees (1971) and Saes (2012)) work that tax setting depends on:

1. The ability of labourers to shift their productive labour hours when taxes changes. Their preference for hard work diminishes as higher taxes are implemented (Frisch elasticities and labour-supply elasticities)

2. The distribution of tax payers (wealth is usually Pareto distributed)

3. Income and substitution effects. One's income is directly affected by taxes but substitution effects depend on labour-supply decisions. Elastic labour supply means that people can change their labour decisions (either choose more leisure or work harder to maintain a similar income as before the tax change). If the elasticity is larger than one, then increasing the tax rate could result in less revenue collection as more people decide to work less (either by reducing the amount of hours worked or by simply not being as productive while working the same amount of hours). When the elasticity is smaller than one then increasing the tax rate could result in higher revenue collection as people work at the same levels as before (same amount of hours and at the same productivity level).

In South Africa one would, a priori, expect labour supply to be very inelastic (smaller than 1) across income earners. Using StatsSA's Labour Force Surveys (LFS) suggests that the labour supply elasticity is close to zero (even while controlling for other costs such as transport or grants received). That means that one could increase tax rates and collect more revenue. But South Africa's top income earners also have access to other income types, which means that they can shift their labour income to capital which has a lower tax rate. I.e. some top income earners have a lower effective tax rate than middle and low income earners which violates the vertical equity principle. A business owner might not get his benefits in terms of a salary (which is subject to labour income tax), but gets his benefits in the form of higher dividends or company profits (which are subject to much lower rates than labour income tax). Thus raising taxes at the top end of the income distribution might only result in tax shifting. One possible policy remedy would be to increase tax rates on capital. Now this comes close to Lump-sum taxation (which implies Pareto-efficiency). Obviously there are different trade-offs involved in taxing capital such as a decrease in investment incentive.

Setting tax rates are notoriously difficult. Just looking at South African data would suggest that it is possible to increase the top tax bracket, or even introduce a new top income bracket. This would not necessarily result in a significant increase in tax burden as the effective rates are still low. This is definitely one way to close the budget deficit without incurring much harm on the economy overall. The only problem is that South African's do not necessarily get a return on their taxes as spending on education and health in the past has yielded little to be happy about (more on this later).