India currently has five consumer price indices (CPIs), three of which are population-segment specific CPIs, namely, CPI-IW (base 2001) for industrial workers, CPI-AL (base 1986-87) for agricultural labour, and CPI-RL (base 1986-87) for rural labour. CPI-IW, covering manual workers engaged in specified industrial sectors, is available at 78 selected centres and at all-India levels while CPIs AL/RL are available in 20 specified states and at all-India levels. As none of these indices provide inflation data for the entire population, the Central Statistics Office (CSO) of the ministry of statistics and programme implementation in 2011 began compiling new CPIs, one for the entire urban population, namely CPI-U, and the other for the entire rural population, i.e. CPI-R. These new indices are compiled at state/Union territory and all-India levels.
Though these indices are produced for different purposes, have their separate commodity basket and weighting diagrams, and a different data collection and compilation mechanism/institution, etc., these indices should be compared. If divergent trends are noticed, they could not only raise the question of reliability but also undermine the credibility of the system of data collection.
Apart from these concerns, which arise with diverging trends, there is a need to see if separate indices are required and whether the benefits of these specific indices outweigh the costs incurred in their compilation.
CPIs AL/RL are used for the fixation of minimum wages of agricultural labourers and rural unskilled wage employees. With a base which is already 30 years old, the question that arises is about the appropriateness of their commodity basket and weighting diagram. The major stakeholders of CPI-IW are the employers and workers, but apart from the base which is already 15 years old, the question that arises is its restricted specific population group and its wider use, which hardly match. A committee of the National Statistical Commission which had looked into the issues of these indices had suggested the use of CPI-U as an appropriate wage indexation method. Even the Sixth Pay Commission for Central government employees found the use of CPI-IW a poor proxy of price rise for the basket of commodities of that group of employees. The labour bureau, which compiles this index and the other two CPIs for agricultural and rural labourers, does not want any deviation from the existing system without wider consultations.
Notwithstanding the turf which the labour bureau may be defending for continued generation of these indices, let us examine whether the new series of CPI Urban and Rural could meet the specific requirements of the population-specific indices.
The two charts (see charts) provide the average annual inflation that these various indices indicate since 2012 at the aggregate level and at the level of commodities belonging to the subgroup food.
Average inflation at the aggregate level of these five indices varies from 7.32% for CPI Urban to 7.99% for CPI-RL, though inter-year inflation differences are quite significant.
All these series, however, appear to be converging. The correlation coefficient for CPI-AL and RL is near 1, indicating that these two series have moved almost together. The correlation of CPI AL/RL with CPI Rural is 0.87, which also suggests that a switchover to CPI Rural for these two indices may not result in any compromises with the specific requirements that these two indices have. The correlation coefficient of CPI Urban with CPI-IW at 0.91 also suggests that this switchover also may not raise any acceptability issue.
We also look at food inflation as reflected by these indices. Average food inflation has varied from 7.31% for CPI-AL to 8.88% for CPI-IW. The correlation coefficient for these indices is almost similar to what has been observed for the inflation at the aggregate level.
The committee constituted by the National Statistical Commission, while suggesting a common agency for the collection of price statistics, also suggested CPI-Rural to substitute for CPI-AL/RL and CPI-Urban to substitute CPI-IW. Monthly movement of inflation based on CPI-Rural and CPI-AL/RL also indicates this would be a wise move.
In fact, CPI-Rural could have much wider use in terms of coverage, with a more recent base and the possibility of being automatically updated with a shorter time lag. The CPI-Urban can be a substitute for CPI-IW.
As in the case of CPI-Rural, CPI-Urban has wider applicability and wider coverage, a more representative commodity basket and an easier base change possibility.
The new series of CPI, together with its rural and urban components, derives its weighting diagram from the National Sample Survey Organization’s consumer expenditure surveys, which are undertaken by the ministry of statistics and programme implementation at regular intervals. Further, these have wider coverage both geographically and sector-wise, in terms of the commodity basket and sample size.
Together with a regular base change, which is near automatic in the new CPI series, a single agency data collection system also provides more homogeneity to the data collection process.
It is time for consolidation and a move over to CPIs Rural and Urban. The necessary adjustments, if any, could be made in the new CPI series to ensure that population-specific requirements are also met.
R. Gopalan and M.C. Singhi are, respectively, former secretary (finance), and former senior adviser, ministry of finance