How many inflations are there?

In 2015, the government notified a new mandate for the Reserve Bank of India (RBI), where it was stated that the objective of monetary policy framework would be primarily to maintain price stability, while keeping in mind the objective of growth. Further, price stability was defined in terms of the year on year change in the monthly Consumer Price Index (CPI) published by the Central Statistics Office (CSO). This decision was also enshrined in the RBI Act in 2016.

The need to maintain price stability as the principal, if not the only, objective of monetary policy is well articulated in the economics literature. But the choice of the instrument is not formally discussed in the same way. In so far as the formal literature in macroeconomics is concerned, the prices mentioned in the discussion of price stability refer to something called the aggregate price level of the economy. If we pursue this further, we can see this price level linked to either the transactional matrix (i.e. the prices underlying all transactions of the economy) or the earnings/income vector (i.e. the prices behind all incomes earned). These different perspectives are sometimes named after the works of Irving Fisher or Milton Friedman.

The CPI is neither. The CPI is an index which measures the prices behind the consumption basket of a representative consumer. If we assume a society with complete equality then this would be the entire consumption basket. It would, however, neglect savings and investments. To give a sense of perspective, in India, of the total national income, approximately 60 per cent is consumed, the rest is either invested or consumed by the government. Thus, the CPI ignores those dimensions of the prices. If we think in transaction terms then CPI at best captures prices of those goods and services which enter final consumption. It ignores the prices underlying the vast structure of intermediate transactions which are inherent in a complex economy. In fact, a feature of economic development has been the increased specialisation of work and activity leading to an increase in intermediation and the further marginalisation of the CPI.
 

If we make additional assumptions about the fixity of the production structure, resource availability and demand conditions then the prices of final goods will be in fixed relationship with the prices of all other intermediates. In such a world, stability in the CPI will automatically ensure or correspond to stability in the overall price level. In a dynamic economy, undergoing rapid technological change, this fixity is a chimera.

Recall further the assumption of perfect equality. In an unequal world, the CPI reflects prices only for a small segment of the population who are close to the average. In a world with changing inequality, the representativeness of CPI itself becomes an issue.

The price index confronting the poor is very different from the price index confronting the rich. To illustrate one facet of this, if you walk into a large department store the prices in the morning are for the most part the same as the prices in the evening. But the neighbourhood or village street market behaves differently as the prices in the morning are very different from the prices in the evening for a large range of perishables.

These examples were merely to illustrate that the link between our measurable index and our theoretical objective of price stability are mediated by a vast set of assumptions on the institutional character of the economy. The choice and suitability of any target are essentially predicated on assumptions of institutional stability. The time horizons in which these assumptions of institutional stability are valid have themselves been shrinking over the years.

These conceptual issues were outlined to illustrate how limited was the discussion behind the choice of a suitable measurable index for the monetary policy objective. There are other problems with the manner in which it is operated.

One example will illustrate the shallowness of the discussions. One of the components of CPI is accommodation services acquired by renting or purchasing housing. The only information readily available is that of rental rates. Even here for approximately 15 per cent of rented dwellings, the rent paid is notional as it is provided by the employer (usually government) to their employees. In these cases, CSO imputes a price by the value of house rent allowance foregone by the employee. Periodically, the government revises the employee’s salary and thus the housing allowance. Whenever this happens, the component of CPI which is based on this calculation changes. It appears as if the rate of inflation has gone up. But this conclusion is similar to measuring your temperature after drinking a cup of tea and noting that the temperature has gone up and concluding that you have fever. Since you would not immediately start a course of antibiotics on the latter discovery, allowing it to influence your monetary policy in the former case is similarly absurd.

The agreement which the government created with the RBI also required the RBI to publish a document on the sources of inflation and its forecasts. The RBI meets this requirement in a procedural or proforma manner, a proper analytical compliance of the spirit of this mandate is still awaited.

The writer is former Chief Statistician of India