Simple solutions to trim inequality do not add up
15 October 2015
Posted by: Author: Gavin Keeton
Author: Gavin Keeton (BDlive)
Much has been written about the need to reduce SA’s exceptionally high level of inequality. It is vital that we address inequality, but few commentators acknowledge how difficult meaningful progress is in achieving this goal.
Proposed solutions usually focus on taxing the rich to spend more on the poor. But it is an arithmetic reality that the more unequal a society, the harder it is to substantially reduce inequality this way.
This is because, in very unequal societies, higher taxes are paid by few people and the increased tax take is therefore relatively low. This increased revenue must then be shared among many poor people, so the effect on each is tiny.
In more equal societies, the number of rich people relative to poor is much higher, so the effects of raising taxes and increasing transfers are greater.
This is confirmed in a recent US study by the Brookings Institute.
The study uses US tax data to examine the effect of raising the top income tax rate on inequality. Scenarios are developed in which the top income tax rate rises from its current 39.6% to 45% and 50%. The model assumes that higher tax rates have no effect on declared incomes.
The additional tax raised under these scenarios is $49bn and $96bn. These are big numbers, but the effect on after-tax inequality is astonishingly small.
The fall in the Gini coefficient measurement of inequality is an almost meaningless about one-fifth of 1% in each case.
A second set of scenarios assumes all the income from the higher taxes is targeted at the poorest 20% US citizens and distributed equally among them.
When the top tax rate is 50%, this transfer adds $2,650 to their annual income. The effect on the Gini coefficient is again paltry.
The study notes that these findings do not necessarily make higher tax rates undesirable. There may be good reasons for wanting to raise tax revenue. But as an instrument for reducing inequality, the "strikingly limited reduction" achieved reveals the limitations of this approach.
Tax data for 2013 suggest that raising the top rates of tax would have similarly small effects on inequality in SA.
In 2013, 8% of assessed taxpayers earned more than R500,000. They earned 34% of all taxable income and paid 54% of all personal income tax at an average tax rate of 31%.
If SA’s top marginal rate was raised to 50%, the average tax rate paid by these top income earners might rise to about 42%, raising R45bn in additional tax revenue. This would increase government revenue by just 5.5%. There would be many competing demands on such revenue, not least reducing the government’s deficit.
But if it was ring-fenced for social grants, it could fund a 40% increase on what was spent on grants to more than 15-million South Africans in 2013.
While increasing grants or extending their coverage may reduce extreme poverty, the effect on income distribution is very small. The amount involved for each poor individual does very little to narrow the income gap. A more substantial effect is achieved only if higher taxes are applied to a much broader band of taxpayers.
The largest number of taxpayers earn between R200,000-R500,000 a year. But just imagine the effect — to say nothing of the political cost — of asking South Africans in these income bands to pay an extra 5% (R10,000-R20,000) in tax each year.
Clearly simplistic solutions are not enough. A range of interventions is needed. These include jobs so that the millions of unemployed South African can earn some income. Improving skills will help people move into higher paid jobs and reduce the premium earned by those whose skills are in short supply.
Achieving this will not be easy. But that is no excuse for pretending simple solutions will work.
- Keeton is with the economics department at Rhodes University
This article first appeared on bdlive.co.za.