to the closures in our quest to understand the impact of electricity price increases. In the short-run,
the capital stock is assumed to be fixed, while the rate of return on capital is allowed to change.
The labour market is modelled in the typical ORANI way, which assumes fixed real wages in the
short-run, and an infinite supply of labour at the given real wage. This is a fairly realistic assumption
of the South African labour market with its large unemployment of unskilled labour. The supply
of land is also assumed inelastic. In the long-run, the real rate of return on capital is fixed, with
capital being allowed to adjust, while employment is fixed, with adjusting real wages.
With reference to the macroeconomic variables, it is assumed that aggregate investment, gov-
ernment consumption and inventories are exogenous, while consumption and the trade balance are
endogenous in the short-run. It differs slightly from the ORANI assumptions of fixed real house-
hold consumption in the short-run, because this specification allows us insight into the effect of the
suggested policies on South Africa’s consumption and competitiveness. In the long-run, we follow
the usual ORANI closure with C, I and G2 endogenous, and the Balance of Trade exogenous3. All
technological change variables and all tax rates are exogenous in the closure. Finally, the nominal
exchange rate is set to be the numeraire in each of the simulations.
The focus of the paper is the impact of electricity prices on the real exchange rate, as measured
by the CPI variable in our model. To shed light on this question, various shocks to the model are
implemented, while altering some key assumptions about employment and consumer behaviour.
4 The Scenarios
Eleven simulations are run to determine what the influence of the following variations in the as-
sumptions would be on the results:
1. the difference between being able to set the price of electricity and raising a tax on electricity;
2. the difference between a tax on households only, versus a tax on intermediate and final use of
electricity;
3. the difference between fixed real consumption and the standard closures;
4. the difference between fixed real wages and fixed nominal wages in the short-run; and
5. the effect of electricity price increases in the long-run.
In variation (i), we compare the situation where the government levies an additional tax of 10
per cent on the price of electricity with the one where the electricity industry increases its own
price by 10 per cent. A CGE model has endogenous prices, i.e., they are determined by the model
and could in general not be “set” by anyone. Usually prices can only be affected through some
exogenous shock to the model, or in other words, through a manual change to some variable that is
not endogenous. Therefore, to be able to set the price of electricity, we need to change the closure
of the model by making the price of electricity exogenous and another variable endogenous. The
only possible way to do this is to endogenise the amount of production taxes paid by electricity, or
some cost variable. The (neo-classical) model does not allow excess profits to any industry, so the
only way that they can actually do this is by either paying more taxes or experiencing higher costs.
Therefore, in principle, it does not matter who increases the price of electricity - the government or
the electricity industry - the effects will be similar.
In variation (ii), we isolate the effect of a tax on only household consumption of electricity. In
general, industries pay much less for electricity than households, and some industries would often
be exempted from price increases. Hence, it is necessary to isolate the effect of a price increase on
2From the well-known macroeconomic equation Y = C + G + I + X- M.
3No country should have continued trade deficits or surpluses in the long-run.