Judging by the recent columns in this newspaper by eminent economists, some of them seem to have been touched by a dose of summer madness. The most egregious is Surjit Bhalla's recent call to send Rakesh Mohan to China to learn how to maintain an undervalued nominal exchange rate.

This is particularly strange coming from the author of a dissenting note to the Tarapore Committee Mark II report on capital account convertibility, who rightly made the case for a speedy and complete opening of the capital account. But if this advice is accepted - as it should be - how can poor Rakesh Mohan maintain an undervalued nominal exchange rate?

All the muddles arise because of the failure to distinguish between three crucially different definitions of the "exchange rate", and the impossibility of the authorities to target the only one, which determines real variables - she real exchange rate.

This is the relative price of two broad classes of commodities produced and consumed in the economy: tradeable goods, whose domestic prices are set by foreign currency prices and the nominal exchange rate - the Rs/$ rate -, and non-traded goods, which because of high transport costs cannot be internationally traded (like housing or haircuts), or which could be traded but have been made non-traded because of import controls.

The third "exchange rate" is the real effective exchange rate, which is based on purchasing power parity, and corrects the nominal exchange rate for the difference between the domestic and foreign price levels.

The REER is only equivalent to the RER (for which it is often erroneously taken as a surrogate) if there are no non-traded goods in the economy. In India, even after the recent trade liberalisations, about 30 per cent of the goods in the WPI are non-traded. The only instrument the authorities can control in the managed floating system current in India is the NER. The RER in particular is an endogenous variable being determined by the general equilibrium effects of capital inflow absorbed, the net effects of monetary and fiscal policy on domestic absorption, and the effects of terms of trade changes.

Given the resulting equilibrium, changes in the NER will not affect the equilibrium RER, but only the money prices of traded and thereby of non-traded goods required for this RER.

The diagram shows the movements in these three exchange rates in India since the 1991 reforms. The REER and NER shown are the 36-country trade- weighted series computed by the RBI. The RER series is one that has been derived with various associates at the NCAER over the years.

It shows first, that the REER does not mirror the RER; second, that since 1996-97, when total foreign inflow (including remittances) surged to over 6 per cent of GDP, the RER has had an upward trend; third, the near constancy of the REER since 2000-01 suggests that it is this rate (which they probably assume mirrors the RER) which the RBI is targeting by requisite movements in the nominal exchange rate and monetary policy.

The relative profitability of tradeable vis-a-vis non-tradeable goods is determined by the RER. If it appreciates (say, as a result of increased absorption of capital inflow) the tradeable sector must necessarily shrink relative to the non-traded sector.

But it can still grow in absolute terms if the economy as a whole is growing rapidly. This growth being more rapid, more foreign savings are allowed to supplement domestic savings, to raise the domestic investment rate.

Despite assertions by many economists to the contrary, there is no special growth- enhancing effect from artificial promotion of tradeables, which can only be brought about by restricting capital inflow, and thence lowering the rate of investment to prevent the RER appreciation which would otherwise occur.

On the monetary side, if the authorities allow the capital inflow in any year to be absorbed, domestic high-powered money will rise by an equivalent amount, which when spent will lead to an equivalent trade deficit and a foreign exchange outflow of an equivalent amount.

The RBI's balance sheet will first show an increase in its foreign exchange assets equal to the inflow matched by an equal increase in its liabilities of increased reserve money. But, as this increase in money is spent, it will lead to a trade deficit equal to the capital inflow, financed by the increased foreign exchange assets the RBI acquired in the first round.

The RBI's balance sheet at the end of this process of absorption will show no net increase in reserve money, nor in its foreign exchange assets. There will be no inflationary impact from the absorption of the capital inflow if the NER is truly flexible. For, the above absorption of the capital inflow will lead to an automatic appreciation of the rupee.

The price level can however rise if the authorities maintain a fixed NER, through appropriate sterilisation operations, allowing only part of the capital inflow to be absorbed. For the absorption of any part of the inflow necessitates a rise in the RER. If the price of tradeables remains constant because the NER is fixed, the only way the RER can rise is through a rise in non-traded goods prices and thence in the price level.

Moreover, if the country faces a deterioration in its terms of trade (with a rise in the foreign prices of its imports relative to exports) as has happened in recent years, ceteris paribus, this will also require an appreciation in the equilibrium RER.

If instead of appreciating the NER it is devalued, as seems to have happened in 2006-07, the price level will rise, even if the increase in money supply is not greater than that required by the capital inflow allowed to be absorbed and the extra money demand generated by a higher growth rate of output.

Thus, rather than any over-expansion of money, the recent rise in the inflation rate could be due to the inappropriate depreciation of the rupee in 2006-07. The tightening of monetary policy recently undertaken would thus only damage growth.

These recent macro-economic outcomes and the muddled discussion that they have engendered once again emphasise that it is now time for India to float the rupee cleanly, confining intervention to some smoothing of the NER. This in turn requires the capital account to be opened fully to allow both economic agents to make their own portfolio decisions and to develop and deepen the necessary forward exchange markets to deal with the inevitable shocks in a dynamic global economy from changing capital flows and terms of trade.