Skip to content

Finding the Optimum Level of Taxes in South Africa: A Balanced Budget Approach

Literature › Works

Yolanda van Heerden; Niek J. Schoeman (2008). Finding the Optimum Level of Taxes in South Africa: A Balanced Budget Approach. University of Pretoria, Department of Economics (unpublished paper dated 29 September 2008).
Type working paper
Scope Post-1994 South African economy
Method econometric
Period studied 1960–2007
Regions South Africa
Themes macro fiscal finance debt, methods data sources, comparative theory, transition post apartheid
Citation key heerden2008

Argument

The paper estimates the growth-maximising ratio of government spending (equal to taxation under a balanced budget) to GDP for South Africa. It uses a Scully-type Cobb-Douglas model with a public and a private sector. The estimated optimum is 21.94 per cent of GDP, well below the 26-28 per cent recorded in 2006-07. The authors conclude that South Africa is on the downward-sloping side of a Laffer-type curve and should cut tax rates to raise growth. The argument rests on OLS cointegration and error-correction estimation of annual 1960-2006 data, with the long-run coefficients adjusted by an Engle-Yoo step.

Evidence and method

Annual South African Reserve Bank series for 1960-2006: GDP, general government consumption and gross fixed capital formation, and the CPI. These are deflated by the CPI and used to build the tax/spending ratio, disposable income and growth rates. Tax-structure ratios for 1981-2007 (Table 1) are also drawn from SARB sources. The international benchmarks come from secondary studies (Scully, Chao and Grubel, Pevcin, Mavrov and others).

Findings

  • The paper's headline result is that the actual tax burden exceeds the growth-maximising level. (p. 1)

    The results indicate that the actual average tax burden far exceeds its optimum level and that the authorities will have to adjust tax policy accordingly in order to improve on the level of economic growth.

  • The estimated optimum is 21.94 per cent of GDP, against realised ratios of 26-28 per cent. (p. 15)

    According to the model the optimum tax rate is 21.94 per cent. The actual level of tax as a share of GDP for 2006, was 26 per cent and in 2007, it increased to close to 28 per cent.

  • The long-run elasticities of growth with respect to government expenditure and disposable income are both positive. (p. 12)

    A 1 per cent increase in government expenditure would lead to a 0.22 per cent increase in economic growth. A 1 per cent increase in disposable income would lead to a 0.77 per cent increase in economic growth.

  • The scatter plot shows most observations to the right of the growth-maximising turning point. (p. 11)

    It is clear that most of the tax ratios over the period lie to the right hand side of the optimum turning point of the curve.

  • The paper links the 1980s sanctions to lost foreign investment and the 1985 debt standstill. (p. 6)

    During the 1980s, the economy endured international sanctions that lead to a loss of international investment and the debt standstill.

  • Government spending rose from about 1970 and then stabilised at a high level. (p. 6)

    After about 1970, the level of government expenditure started to increase and fluctuated at an average of 26 per cent between 1985 and 2007.

  • The composition of tax revenue shifted after 1994, with personal income tax falling as a share of the total. (p. 6)

    Personal income tax comprised 42 per cent of total tax revenue in 2000 but its share declined to 30 per cent in 2007.

  • Post-1994 corporate and top personal tax rates were cut substantially. (p. 6)

    The corporate tax rates decreased from 40 per cent to 28 per cent and the top marginal personal income tax rate from 45 per cent to 40 per cent

  • The post-1994 tax burden was officially described as stabilised at about 25 per cent of GDP (Finance Minister, 2002). (p. 6)

    stabilisation of the tax burden at approximately 25 per cent of GDP

  • A 1994 dummy was used to control for a structural break from the political change. (p. 21)

    A dummy variable was incorporated to account for the structural break caused by the political change in 1994 which distorted the available time series.