Towards a better framework

My column last month on inflation and monetary policy provoked more response than usual, no doubt because of the heatedness of the debate, rather than any one thing I wrote. With some of the heat having subsided, this is a good time to revisit some of the issues. This is important since the long run goal of efficient monetary management remains, in my view, unrealised.

One indicator of sub-optimality is the range of opinions that were expressed over the past few weeks, on the conduct of Indian monetary policy. Some divergence of views can be expected based on different interests, but there was even disagreement on the basic facts of how much monetary tightening had occurred, let alone what was optimal. One conclusion I would draw from the past few weeks is the need for greater transparency and predictability of monetary policy. Predictability includes timing, but also can have implications for the range of instruments that the RBI uses. A more parsimonious use of instruments could promote simplicity, and hence the predictability of what any policy moves would accomplish.

Last month, I raised the idea of using the Taylor rule framework to assess the RBI?s monetary policy stance. One former policy advisor suggested to me that I was using the term ?inflation targeting? too narrowly, and that a high enough inflation coefficient in the Taylor rule would constitute inflation targeting. The issue remains as to whether the RBI?s record fits the bill. I was pointed to the work of Vineet Virmani, of IIM-Ahmedabad, who estimates monetary policy rules for India for the period 1992-2001. A striking feature of Virmani?s results is that the estimated rule is extremely sensitive to the inflation measure used. Using the headline measure of inflation, the WPI suggests that the RBI?s ?rule? (as implied by the data) has been destabilising, while an alternative (technically, a ?trimmed mean?) indicates a rule with much better inflation-fighting properties (inflation-term coefficients of 1.6 to 2, versus 1.5 in Taylor?s original rule). Clearly, there is work to be done here in deciding what to measure and target. Virmani also allows for gradual adjustment of the interest rate, and suggests that the RBI?s behaviour fits a different ?McCallum rule,? in which nominal income is targeted.

One aspect of the RBI?s conduct of monetary policy is its continued use of the cash reserve ratio (CRR). In the 1990s, when the external and internal challenges were greater, the CRR was used regularly. In January 2002, RBI Governor YV Reddy stated plainly that, ?The medium-term objective is to bring down the CRR to its statutory minimum level of 3.0% within a short period of time.? Clearly that has not happened so far, and raising the CRR has been an important tool for the RBI in the last year. The RBI also continues to use provisioning requirements for standard advances in specific sectors, such as real estate. These instruments show up indirectly in empirical estimates of policy rules. For example, Virmani finds evidence that monetary aggregates are still implicitly targeted, in his estimated rule. It was these kinds of direct interventions that I implicitly characterised last month as ?command and control.? Perhaps that language was too strong, as a senior policymaker suggested to me. However, such instruments make the guiding principles and impacts of monetary policy less clear. It would be better, in my view, to move away from such instruments (which can seem too reactive) toward more forward-looking interest rate management.

A final issue that has been well discussed in the media is the RBI?s exchange rate management. An exchange rate target clearly shows up in Virmani?s estimates of the RBI?s implicit policy rule. We know from theory that fixing the exchange rate with free capital flows cedes control of domestic monetary policy. The RBI is grappling heroically with the so-called ?impossible trinity.? The rupee has been allowed to appreciate against the dollar, and some additional capital outflows allowed. It remains to be seen how this will play out. Economic orthodoxy would suggest that India is now on the right path, with gradual capital account liberalisation, a freer exchange rate, and potentially better control of domestic monetary conditions. There is a minority opinion, however, which views with favour the Chinese model of pegging the currency to support exports, maintaining capital controls, and allowing foreign exchange reserves to pile up. Some of the difference in these perspectives depends on expectations of how quickly and strongly the Indian financial sector will develop.

In either case, there seems to be room for the RBI to move toward a monetary policy framework that uses fewer instruments, strives to be more forward-looking, and commits to simple, transparent policy rules. One can infer some of this direction from recent policy pronouncements, but perhaps not enough. It also may be the case that a new contractual relationship between the government and the RBI is necessary (implementing inflation-targeting contracts, sometimes called ?Walsh contracts,? after my colleague Carl Walsh, who pioneered the concept). But progress in policy making can be made even without that institutional innovation. It is easy to second-guess every small policy move made by the RBI. I am not doing anything like that here, or in my previous column. The goal I am pushing for is a move towards a superior policy rule or framework.

?Nirvikar Singh is professor of economics at the University of California, Santa Cruz