It is possible to capture the dynamics of the operational conduct of current monetary policy through a small, computable model linking the objectives and instruments of monetary policy within a simple representation of the macroeconomic processes through which monetary policy is transmitted. The model is set in the modern tradition of the 'consensus' approach which is being increasingly employed to capture the prevailing operating procedures of central banks worldwide. The 'consensus' model has the advantage of being simple, transparent and tractable. Relative to large-scale macroeconomic models, the consensus approach allows for easier evaluation of alternative policy rules. Although highly aggregative, these models have considerable theoretical content in providing stylised representation of the monetary policy transmission mechanism.

The model estimated in this Chapter draws from the empirical approach adopted in the context of examining the impact of bond financing of fiscal deficits in India (Moorthy, Singh and Dhal, 2000) and the foregoing discussion: i) monetary policy is faced with a multiplicity of objectives, although a single goal is desirable; specifically, monetary policy cannot avoid concerns about stabilising the real economy around a desired growth path even as it commits itself to price stability; ii) there exists a short run trade-off between growth and inflation and monetary authorities are constrained to take a position on the growth-inflation curve, reflecting their policy priorities; iii) money plays no explicit role in the model although the key monetarist assumption that inflation in the long run is determined by monetary policy is retained implicitly; iv) the interest rate is treated as the principal instrument of monetary policy in consonance with monetary policy operating procedures all around the world under which monetary authorities adjust interest rates in response to economic developments. The interest rate becomes an overall index of financial conditions and all financial prices move in a stable and predictable way with changes in the policy rate (Meyer, 2001).

The first issue is the choice of objective(s) for monetary policy. The stated objectives of monetary policy in India are to 'provide sufficient credit for growth while ensuring that there is no emergence of inflationary pressures on this account' (RBI, 2001). These objectives are consistent with announced projections of growth and inflation, with projections for non-food credit performing the ad hoc role of an intermediate variable. The choice of these objectives inevitably involves a consideration of societal welfare, i.e., deviations (of inflation and growth) from the chosen combination cause losses of macroeconomic welfare. There is considerable agreement among academics and central bankers that the appropriate loss function involves stabilising inflation around an inflation target as well as stabilising the real economy represented by the output gap (Svensson, op cit). The conduct of monetary policy should be directed towards minimising these welfare losses. The critical policy choice is the relative importance or weights to be assigned to deviations of output and inflation from the targets. This involves a knowledge of the structural characteristics of the economy. In terms of the model, this can be expressed through a simple three equation system specifying aggregate demand, aggregate supply and a policy rule setting out the response of the monetary policy to fluctuations in demand and supply.

As in the case of fiscal policy, monetary policy operates by influencing the gap between aggregate demand and supply which, as indicated in Chapter IV, has to be estimated within the framework of a macroeconomic model (as in Chapter VIII). For operational purposes, however, a reasonable approximation of the aggregate demand-aggregate supply gap can be obtained as the difference between actual and the underlying or structural component of output. Therefore, the first building block of the model is the specification of the behaviour of aggregate demand (output gap) that depends upon its past behaviour and the real interest rate which has a negative impact on demand. The real interest rate is defined in a forward-looking sense by deducting inflation expectations (proxied by lagged inflation) from the nominal interest rate represented by the SBI advance rate. In India, cyclical changes in economic activity have often been induced by supply shocks, predominant among which has been that emanating from increases in international oil prices. This is explicitly incorporated in the estimation of the aggregate demand in the model in the form of fuel price inflation. Essentially, in terms of standard economic analysis, the aggregate demand function takes the form of an expectational IS (Investment-Saving) function of the type that can be justified by dynamic optimisation analysis.

Deviations of inflation from the threshold can be studied from the supply side of the economy. The aggregate supply curve can be represented by a Phillips curve type formulation in which inflation is positively related to the output gap. The supply curve also captures the effect of inflation expectations on inflation with expectations formed in an adaptive way, i.e., current inflation is determined by its past. The introduction of lagged inflation in the estimation of the supply curve produces the short-run tradeoff between inflation and growth. Food prices, another major source of supply shocks, are introduced into the supply curve reflecting the structural characteristics of the Indian economy.

The model is closed by specifying a policy rule setting out the response of the monetary policy to deviations of the economy from the desired combination of growth and inflation chosen by the monetary authority to minimise welfare losses. In view of the country-specific situation involving the conduct of public debt management by the monetary authority and the recourse of the fisc to money financing, the fiscal deficit is explicitly introduced into the monetary policy reaction function. The policy reaction is embodied in the adjustments in the Bank Rate, the principal signaling variable as well as an anchor for interest rates on standing refinance facilities, repo rates and in general, the term structure of interest rates in India. The deployment of the Bank Rate is postulated as being symmetric in response to inflation and output deviations. An increase in the Bank Rate results in a rise of the market rate for advances, which compresses aggregate demand (or output gap) and aligns the output/inflation deviations with targets/threshold.

The statistical results of the model turn out to be reasonably robust1. In the aggregate demand equation, the coefficient of aggregate demand with respect to real bank lending rate is estimated at (-)0.24 in the short-run and almost (-)0.5 in the long-run. This suggests that a 100 basis point reduction in the real interest rate raises the aggregate demand and narrows the output gap by almost 25 basis points in the short-run and up to 50 basis points over time. Supply shocks embodied in the international oil price hikes have a major influence on aggregate demand, generating cyclicity in output behaviour. Adjustments of aggregate demand fluctuations to its underlying levels are fairly rapid, with almost one-half of the output gap closed within one year. This suggests that the amplitude of the cycle in the Indian economy is relatively small - upturns quickly follow downturns, vindicating the results of the spectral analysis conducted in Chapter III. The operation of monetary policy through real interest rates appears to have had a stabilising influence on aggregate demand. Simulations of the estimated equation indicate that the impact of policy measures for demand compression during the balance of payments shocks of 1990 had prolonged effects. Aggregate demand was constrained well below its potential up to 1996-97 when, following the cyclical catch-up, aggregate demand was predicted to experience a downturn. The behaviour of actual aggregate demand indicates that the model overpredicted the downturn, mainly on account of demand stimulus emanating from the Fifth Pay Commission award and a strong rebound in agricultural.

The estimates of the aggregate supply equation2 indicate that the dynamic or long-run effect of aggregate demand on (changes in) inflation is about 0.5, i.e., a 100 basis points increase in the aggregate demand (above its underlying) results in almost 50 basis points rise in the inflation rate. The role of food prices, a traditional channel for transmission of supply shocks to domestic inflation, is also significant; an increase of 100 basis points in foodgrains inflation rate leads to an increase of 44 basis points in the inflation rate (Chart V.7). The estimated supply equation captures the turning points fairly closely.

Estimates of the monetary policy reaction function satisfy a priori expectations of symmetrical response of the monetary policy to output and inflation movements3. An expansion of the aggregate demand and/or rise in inflation above threshold provokes a revision in the Bank Rate in the similar direction. The estimates show that the Bank Rate is positively related to the size of the fiscal deficit (the long-run coefficient of fiscal deficit/GDP ratio is placed at 1.4) indicating that the monetary policy reaction is partly attributable to the size of the fiscal deficit which imparts a downward rigidity to the interest rates in the country. The lagged coefficient of the Bank Rate turns out to be significant suggesting a strong order of interest rate smoothing in support of financial stability, a result consistent with the empirical evidence emanating from the estimated reaction functions of other major central banks (Clarida, Gali and Gertler, 1998).

As real interest rates employed in the model are derived from the commercial bank lending rate (SBI Advance Rate), the latter is linked through a linear relationship with the Bank Rate

In the model, a macroeconomic disturbance resulting in a demand-supply imbalance provokes a monetary policy reaction in the form of a change in the Bank Rate. The commercial bank lending rate changes in a linear and symmetric manner. As inflation expectations are inertial in the short-run, the change in nominal interest rates is reflected in real interest rates. This has a negative impact on aggregate demand and, in turn, on inflation. The monetary authority analyses the developments in aggregate demand and inflation in response to its initial interest rate move and re-calibrates, if necessary, its reaction. A negative shock to ( reduction in) the Bank Rate, as discussed in Chapter VIII, produces some positive effects on the real economy in the short-run, but has adverse effects on inflation over time. In the long-run, therefore, monetary policy has to constantly rebalance the short-run output gains against the loss of welfare due to higher future inflation. This is the constraint facing growth-oriented monetary policy.

Despite the recent record of low inflation the world over, the current disinflation cycle affecting both developed and developing countries has generated considerable scepticism about the primacy, and, in some cases, the single mindedness of monetary policy pursuing inflation. In recent years, there has been a proliferation of empirical work on measuring the macroeconomic costs of reduction in inflation from a certain low level. As Japan has learned, and Europe may soon find out, there is a new danger - falling prices may lock countries into a spiral of economic decline (The Economist, 1999). Given the short-run trade-off between inflation and output on account of wage and price rigidities, it stands to reason to expect a point beyond which attempts to reduce inflation would necessitate a reduction in output. In other words, a sacrifice of output is inevitable in the pursuit of inflation reduction. This loss of output, referred to in the literature as the sacrifice ratio, has encouraged the use of several alternative methodologies for their measurement, largely in the tradition of Okun. In the context of an increasing mass of developed economies having already achieved a relatively low inflation in the 1990s, the concept of sacrifice ratio, and the appropriate conduct of monetary policy, is being subjected to critical review. Of particular concern is the increasing evidence obtained from estimation of sacrifice ratios for industrial countries that aggregate supply curves tend to flatten out in an era of low inflation and structural rigidities. This may increase sacrifice ratios and lead to higher unemployment as firms would be forced to adjust, not through changes in wages which acquire an additional rigidity at low inflation rates, but through changes in employment. Thus, a tightening of monetary policy would have stronger real effects than in the past. The question, therefore, is raised as to whether "gain from reducing inflation to zero is worth the sacrifice in output and employment that would be required to achieve it".

The empirical evidence points to rising sacrifice ratios in periods of declining inflation. For 19 industrial countries for which inflation rates had fallen from 8 per cent (1965-85) to 3.5 per cent (1985-98), an increase of almost 75 per cent in the sacrifice ratios from 1.5 to 2.5 was observed (Anderson and Wascher, 1999). At the same time, the size of sacrifice ratio also depends upon the speed of adjustment with gradual disinflation leading to a higher sacrifice ratio compared to a cold turkey approach providing empirical support for rapid disinflation (Ball, 1993). The preference for speed of disinflation would critically depend upon the shape of the social welfare function. The increase in sacrifice ratios is also found to be relatively less for the countries that had undertaken measures to increase labour market flexibility.

An important related issue is whether increased central bank independence lowers the sacrifice ratio, i.e., whether there is a credibility bonus associated with independent central banks? A priori the public is expected to more readily believe the anti-inflationary pronouncements of an independent central bank and this, in turn, should lower output losses. Some studies, however, did not find any such credibility bonus; on the contrary, a positive and significant relationship between central bank independence and sacrifice ratio was observed. In other words, more independent central banks, on an average, pay a higher output price per percentage point of inflation to reduce the inflation rate (Fischer, 1994). The results suggest that even independent central banks have to fight hard and long to bring inflation down after an inflationary shock has struck.

Estimations of sacrifice ratios for developing countries, and particularly, for India are scanty. Nevertheless, given the current low level of inflation, concerns have been raised regarding the efficiency gains, and, more recently, on the potential output losses associated with assigning relatively higher weights in the conduct of monetary policy to reduce inflation. Accordingly, an attempt is made to estimate the sacrifice ratio for India from the aggregate supply curve (Phillips Curve) estimated as an element of the operating framework of the Indian monetary policy (presented earlier in this section). The specification of the inflation process in the aggregate supply curve follows the 'triangle model of inflation' (Gordon, 1997; Turner and Seghezza, 1999). The sacrifice ratio5 for India turns out to be almost 2 over the period 1975-2000, i.e., in a low inflation environment, a further reduction in the inflation rate by 1 percentage point would reduce the output by 2 percentage points from its potential level.

The estimated sacrifice ratio for India is relatively lower than that of major industrial countries. Estimates of sacrifice ratios are found to be sensitive to the estimation methodology. For a sample of 19 countries, the estimated sacrifice ratio averaged 5.8 per cent based on quarterly data and 3.1 per cent based on annual data (Ball, 1993). Similarly, for OECD countries, based on single equation estimates, the average sacrifice ratio was 3.2; for most of the countries, the ratio lay in a range of 2-4, although outliers of 1.6 (Japan, Italy and the Netherlands) and 7 (Norway) were also observed(Table 5.3). System estimates yielded a common sacrifice ratio of 3.7 for 15 out of the 17 sample countries (Turner and Seghezza, 1999). Given that the threshold inflation in India is at around 5 per cent and that the average inflation has been higher than the threshold, the lower order of sacrifice ratio for India compared with the OECD countries is to be expected; however, once the inflation falls below the threshold inflation of 5 per cent, the sacrifice ratio could end up being higher.