CHOICE OF THE BASE YEAR
The base year of a price index series has to be carefully chosen since the base year would have a considerable influence both on the movement of price relatives of the individual commodities and on the weighting pattern thus influencing the movement of the Index as a whole. The considerations that would go into the selection of the base year could be briefly listed as follows:
4.2 Keeping in view the above criteria, the Working Group examined data relating to production in different sectors of the economy as well as data on prices and several other related indices for the years 1989-90, 1990-91, 1991-92 and 1992-93. The choice of the base year was indeed quite a difficult exercise considering not just the criteria noted above but also the rapid developments which have taken place in the Indian economy as well as in the policy framework within which the economy functions.
4.3 When the Working Group initially deliberated on the choice of the base year, it turned out that detailed results of the Annual Survey of Industries (ASI) 1989-90 were still not available and were expected to be made available within a year hence. It was therefore believed that it might not be feasible to go beyond 1989-90 in selecting the base year, no matter how the profiles of the subsequent years turned out in terms of the criteria laid down. Of course, some controlled information were available for later years for their utilisation in constructing the weighting diagram. It was also found that the deceleration in the industrial production in the year 1992-93 rendered that year unsuitable for selection as base year. There was a great deal of discomfiture that the choice of the base year would be a year in the decade of the 1980s when in fact the indices were expected to be available for use only in the latter part of the 1990s and would be in use mostly in the new millennium. There was also some concern that some major policy initiatives which had taken place in the years beginning with 1991-92 could not be properly reflected in the new series anchored in the year 1989-90.
4.4 Considering all these, and the fact that data on production and prices of primary commodities would be available for more recent years, it was felt that an attempt should be made to consider a subsequent year as the base year. It was also pointed out by the Sub-group on Analytical Issues, that a large part of the economy belonging to the services sector, a fast changing face of the Indian economy, was being left out of the purview of the ‘commodities-only’ wholesale price index. The concept of the wholesale price, the underlying changes in the economy and the coverage of the series were all up for reconsideration, and the choice of the base year itself could not be left out of focus. It was felt that an attempt should seriously be made to consider a later year to be the base for the weighting diagram.
4.5 All this was deliberated towards the end of 1994, and when a lot of subsequent developments were yet to reveal themselves both in the evaluation of the Indian economy as well as the finalisation of the Working Group report, it was decided that work should proceed on the assumption that detailed ASI data for the year 1989-90 should be available around March 1995 and hence that year could be considered as the base year.
4.6 In April 1995 Shovan Ray and D.N. Chaturvedi of the Sub-group on Analytical Issues came up with an alternative methodology for the choice of the base year and construction of the weighting diagram by a method that they termed as "stratified compartmentalised system of weighting". It was a combination of the "top-down" approach discussed in chapter 6 and adopted as the methodology of the weighting diagram for this report by using aggregate data for a recent year and detailed weight shares for the individual items of the distant past. This methodology was an attempt at that time to overcome the inadequacy of the data base and at the same time to extract the best possible information that was available till date. The proposed methodology by Ray and Chaturvedi is enclosed with the report of the Sub-group on Analytical Issues as such eventualities may recur in future. In fact its possible future use in one situation is clearly envisaged and discussed in chapter-7. It is proposed therein that the base year may be changed more frequently, perhaps twice in a decade. The choice of a desirable base year as per the criteria listed above may make it imperative to use the methodology from time to time. The alternative methodology was endorsed by the Working Group for adoption in the choice of the base year and the construction of the weighting diagram alongside.
4.7 However, more hopeful signs on the availability of better data seemed to appear on the horizon. It transpired over time that it might not be necessary to fall back on the alternative methodology of "stratified compartmentalised system of weighting" since detailed ASI data for manufacturing sector were to be made available for the year 1993-94. Data on primary markets were also available for that year. And so 1993-94 seemed to suggest itself as the natural choice for the base year, as it was considered a normal year and was also adopted as the base year for some other related index series at the national and the disaggregated levels. And it appeared then that the National Sample Survey Organisation (NSSO) were in a position to provide disaggregated data for the same year with respect to the unorganised/unregistered manufacturing sector through their survey of the industries in the Directory Manufacturing Enterprises (DME), Non-Directory Manufacturing Enterprises and Owned Account Enterprises (OAE). It was therefore natural that the year 1993-94 would be chosen as the base year; and so it was.
4.8 It turned out subsequently however that the promised disaggregated data from the NSSO on DME, NDME and OAE were not to be available due to some significant snags in the collection of scientific data. The Central Statistical Organisation (CSO) did come out with detailed data at the four-digit level in their Annual Survey of Industries for 1993-94 in 1997 and a new series on the Index of Industrial Production (IIP) was published in July 1998 with the year 1993-94 as their new base year. In May 1999, the CSO followed through with the new series on National Accounts Statistics with the base year 1993-94. With the choice of the base year 1993-94 for the revised series of the WPI, there seems to be greater congruence of various statistical series of economic data that was desired by the Technical Advisory Committee but which could not be fulfilled to any desirable degree until recently.
4.9 As is well known, the choice of the base year has considerable statistical significance through its impact on the weighting diagram. However, no such analytical significance is attached to the "reference year", which is more to do with comparability and compatibility with other indices which are used in the country and elsewhere. The reference year purports only to change the index series proportionally for the base year and, therefore, does not affect the statistical properties of the original base in any way. The Working Group recommends that 1993-94 will be both the "reference" as well as the "base" year for the revised WPI Numbers.