Taxation, Risk and Real Interest Rates

It is an established orthodoxy that negative real interest rates constrain development, by causing financial repression.' The World Bank and the International Monetary Fund (IMF) frequently make the raising of nominal interest rates above the rate of inflation a condition of their lending; and some countries without IMF or World Bank programmes have announced that positive real interest rates are a policy objective. This paper does not challenge the argument that negative real interest rates should be made positive. Rather it raises two frequently neglected points: - the effect of tax rules on whether a positive real rate of interest is actually achieved and how it should be measured' - the risk attached to borrowing at high nominal rates of interest, even though inflation may also be as high or higher. Firstly (Section 2), if the interest paid by borrowers is treated as a tax deductible expense, then it is not sufficient to raise the nominal rate of interest above the rate of inflation in order to have a positive real cost of borrowing. Similarly, if interest received on bank deposits and other financial instruments is taxable, the real return is not necessarily positive even if the nominal rate of interest is above the rate of inflation. Moreover, allowing for tax drives a gap between the nominal interest rates required to make real rates positive for taxpayers on the one hand and non-taxpayers on the other hand. This gap increases with the level of inflation. The nominal interest rates required, to make real after-tax rates positive, are higher the higher is inflation and the higher is the marginal tax rate. As a result, in order to have positive real interest rates after allowing for tax, nominal rates have to be at levels that are likely to be unacceptably high. Only if inflation is low, are tax-adjusted real interest rates likely to be positive. The effect of taxation on the actual cost of borrowing is demonstrated in Section 2 for Sub-Saharan Africa, where a majority of governments have a positive real interest rate policy, some as part of IMF conditionality, some independently. The point also applies in developed countries, but the effect of taxation is smaller where inflation is low, as is shown in Section 2 for the United States. Secondly (Section 3), the higher the rate of inflation and therefore the nominal rate of interest required to achieve a positive real interest rate, the greater is the risk to the borrower. In other words, paying 55 per cent to borrow money when inflation is 50 per cent is much riskier than borrowing at 5 per cent when there is no inflation. The difference between interest rates and inflation may increase; and output prices may not rise in line with inflation, while input prices almost certainly will. This may explain why in countries with high rates of inflation, nominal rates of interest so often remain below inflation, even when the government has announced a policy of moving to positive real interest rates. This point is demonstrated statistically for the countries of anglophone Africa. The conclusion (Section 4) must be that reducing the rate of inflation is preferable to raising nominal rates of interest as a way of achieving positive real rates of interest. The risks and distortions which arise from having a nominal rate of interest higher than inflation can only be minimised by reducing inflation, or by changing the tax laws so that interest is not tax deductible. Where inflation does remain high, it is not possible to say whether the cost of high nominal interest rates is greater or less than the cost of negative real rates. There is nothing new or unorthodox about arguing that it is better to achieve positive real rates of interest by lowering inflation than by raising the nominal interest rate [Killick and Martin, 1990]. This paper points out the difficulty created by tax rules in achieving positive real interest rates, and the risks attached to high nominal interest rates as a way of making real interest rates positive.