This working Group addressed analytical and conceptual issues by casting a distinctive focus on them. It was felt that while issues of weighting diagram, choice of markets and specifications, and periodic revisions of the base year values were essential elements in the revision exercise, they were not exhaustive enough to address the issues that are emerging in an economy changing at a pace and in variety that is qualitatively different from those which had occurred hitherto. Even the concept of what may constitute wholesale prices and the defining boundaries of their reaches were open to various interpretations and subject to scrutinity, not just in India, but in other parts of the world as well. With the dismantling of the controlled sectors of the economy to a considerable degree now and a process that is expected to increase in future, even the methods of data collection and the stages at which such data might be available, became subjects for analysis. The issues which became matters for record and figured in animated discussions are presented below.

7.2 There is no change in the concept of wholesale price or in the system of price collection between the existing and the revised series. As regards the nature of price quotations used in the compilation of the wholesale price index, they represent bulk transactions, generally at an early stage of trading. The prices pertaining to bulk transactions may have divergent nomenclatures, e.g., farm harvest prices, wholesale prices, procurement prices, support prices, administered prices, import prices, export prices, forward prices etc. This has been so over several decades in the past. However, there is a hint of a major conceptual change on the horizon.

7.3 Wholesale prices have divergent connotations as determined by their use by different departments. The Ministry of Agriculture has defined wholesale price as the rate at which a relatively large transaction of purchase, usually for further sale, is effected. According to the Office of the Economic Adviser, Ministry of Industry which compiles the wholesale price index, wholesale prices represent transactions at the primary stage which broadly correspond to producers’ prices. In actual practice, however, the primary sales are not always held ex-farm, ex-mines or ex-factory. Sometimes the sales are held ex-warehouses and in such cases the price quotations may include an element of the cost of transportation from the farm, mine or factory to the warehouse. In some others, the terms of delivery may include the cost of transportation up to the point of consumption. This connotation confines wholesale prices to those corresponding to commodities traded in the economy.

7.4 Two issues are pertinent to this discussion of the concept of wholesale prices. First, for many commodities a wholesale trading market may not exist. For example, for large capital equipment, parts of infrastructure projects or even for consumer durables such as passenger cars, purchases are made on specific project basis or on retail or final consumption basis. However, there is no conceptual difficulty in including them in the series on price quotations because the characteristics of reliability and relative stability of price quotations in time and space is the hallmark of the wholesale trade market.

7.5 The second issue noted here has been important in the past and will become more pertinent in future as the economy moves towards market based competition. In the past, a whole array of prices, particularly of manufactured products, which dominate the weighting diagram, could be collected directly from producers. When market based competition was limited both internally and by tariff protection and government-controlled and administered prices ruled the day, it was possible to fetch price quotations in somewhat matter-of-fact way. As competition is opened up, pricing strategies based on cost information and cost paring assume paramount importance; and as the office of the Economic Adviser, Ministry of Industry, has already begun to experience, obtaining producer prices with the willing cooperation of producers would become less and less the rule, and more an exception. There would then be the imperative to obtain data directly from market trades, presumably at the wholesale stage, which are reliable and relatively stable in time and space. Prices obtained from markets would also be more representative of the real "wholesale prices".

7.6 The Working Group considered whether there is scope for widening the boundaries into non-commodity items in the services sector for inclusion in the weighing structure of the wholesale price index series. Conceptual problem arise in considering such an option. While in some segments of financial instruments wholesale trade may very well take place, in others such as personal services, a wholesale market may not exist. This issue is discussed further below but the Working Group was seriously engaged in considering the scope for its inclusion.

7.7 Service sector constitutes nearly one-half of the economy and is growing in importance. We would recommended that in the first instance, a separate exercise may be undertaken to develop a price index for services sector. This in itself will be a challenging task.

7.8 The Wholesale Price Index series as traditionally constructed is based on the well established Laspeyre’s Index, which is an approximation to the "ideal" index propounded by Irving Fisher. The index uses quantity weights, all of which are expressed in (the homogenous) value measure by converting them to traded value shares of the product by means of their associated prices. So long as these weights are approximately correct during their life time, and if they do not move in any systematic fashion with their corresponding prices, this approximation is a good one in view of the enormous gain in the economy of data collection effort as compared with either Fisher’s ideal or other methods. It is, therefore, best that weighting diagrams are restructured from time to time to correct for the distortions that build up over time. This is the very rationale for setting up the Working Group once about every decade.

7.9 It has been observed that during the recent years the structure of the Indian economy – and indeed the world economy - has been changing at a pace which is much faster than what we have experienced since the WPI series was first brought into place. The reasons are several, mostly arising out of technology revolutions at accelerated pace, but they need not distract attention here. It is, therefore, apprehended that the weighting diagram would get out of tune with its underlying structure sooner than it has been in the past. It is also the case that many dynamic sectors in the economy, including some manufactures, various financial services, information technology, transport, communications etc., have developed in such a way that their prices have strongly influenced their market size (and therefore their traded value share) and hence a strong correlation would be observed between prices and current weights for a wide variety of sectors in the economy. This second factor will also make the Laspeyre’s index numbers lose their desirable property of being proximate to the "ideal" one. There is also a third factor relating to the volatility of the world economy that could distort the index number greatly. For instance, changes in exchange rates over a period of five to ten years could build up to an enormous difference with their base year rates and affect the prices of importables considerably. Consider for instance imports of (crude) petroleum, which is forecast to be imported to the extent of more than half its total domestic use in the next couple of years. Its effect on the weighting diagram will work through essentially three different routes. First, through its own international price changes. Secondly, the exchange rates that are associated with the consignments imported. Thirdly, its effect through its characteristic as a universal intermediate, i.e. entering into the production process of virtually all products and services either directly or indirectly. These three factors combined would considerably affect the expenditure shares of various commodities and services which would form the basis for the weighting diagram for the base year. What is critical, all these changes will be systematic and significant within a short period of time. It is worth noting that out of the three factors stated for the example given above, the first two are comparatively of recent origin and were not of much concern till about the mid-70’s. In future, parametric changes in the world economic variables could not be ignored especially when integration with the global economy is a matter of consensus among economists and policy makers.

7.10 The Working Group, therefore, recommends that the base years, and along with these the weighting diagrams, were changed perhaps twice rather than once every decade. It fortunately happens that a somewhat more frequent revision in the weighting diagram could be made possible due to the enormous strides that information technology industry have made. The transmission and processing of data can now be made with truly enormous speed that could not have been visualised even a decade before. It may thus be possible to undertake such an exercise about twice as frequently.

7.11 If such a course of action is adopted, it is conceivable that a practical problem may arise in constructing the weighting diagram every five years or so. This is because detailed ASI estimates are unlikely to be available for every year, and the year with detailed data on manufactured products may not meet with all the criteria listed in chapter-4 in choosing the base year. In such an event, it is possible to overcome the practical difficulty by constructing the weighting diagram in the manner described as "stratified compartmentalised system of weighting" in chapter-4.

7.12 It may be parenthetically noted that in a separate exercise relating to the construction of the weighting diagram, it has been decided that petroleum crude prices would not form a part of the revised index. Its weight has however been distributed pro rata among its final products.

7.13 The Working Group has also recommended that a Standing Committee be set up to take timely action on issues of conceptual nature and practical difficulties as they arise after the present Working Group is disbanded. The proposed Standing Committee may consist of officials and experts in this and related areas. It should devise ways of incorporating minor changes in the basket when the situation becomes imperative in the data collection process, but consistent with the main principles of the weighting diagram as finalised by this Working Group. The proposed Standing Committee would also be able to guide the work relating to the continued development of price indices for the services sector and attempts to stabilise and merge them into the mainstream of wholesale price indices. It is recommended that the proposed Standing Committee may come into being on the expiry of the current tenure of the Working Group and may be subsummed in the Working Group when such a Working Group is set up to undertake comprehensive revision in the series. The draft terms of reference of the Committee is enclosed as Annexure-V of the report.

7.14 The present system of data collection through mail seems to be quite passive due to the significant proportion of non-response and delay in sending the data, thus necessitating the preparation of provisional index by repeating the old data. The final index, when prepared on receipt of required data frequently differ substantially from the provisional index and thus creates controversies. Hence, it is felt that there should be a direct system of data collection to eliminate such problems. It is suggested that attempt should be made to set up a network of data collection system with the help of NIC, and data should be sourced from wholesale markets as far as possible.

7.15 The Working Group has underscored the importance of including the services sector in the Indian economy in the WPI structure. While estimates of its contribution to the GDP are tentative and largely depend on how one defines the "services", certain estimates claim that its contribution in the base year of the current series (1981-82) was as much as 35 percent even though the sector could not be incorporated in the WPI measure. In 1995-96, the same share apparently increased to 42 per cent and, as the available trends indicate, its contribution is inching up to claim nearly a half of the country’s GDP. The sector is growing at a faster rate in comparison to other sectors. Both conceptual problems as well as practical difficulties have managed to prevent its inclusion in the WPI. There is also some scepticism, no doubt, about the desirability of its inclusion.

7.16 However this extension of the concept of the basket of wholesale prices, to include services along with commodities, is a significant conceptual leap and hence should be adopted with due circumspection and care. The Working Group has taken the view that the WPI index in reality should measure the value of money in the economy as a whole, and, therefore, all transactions that take place therein should be included in this measure. To fix ideas, in the celebrated macroeconomic identity MV = PT, the P that is represented therein should correspond to the price level as measured by the WPI for gross total transactions. This is because the prices refer to the level of activity in the economy and measure the value of money in terms of the gross value of transactions, irrespective of the nature and the complexity of its stratified structure. Money is the numeraire in the system and the price index is the measure of the value of money in terms of all other activity as a composite whole.

7.17 If services are included in the WPI, the modified measure should look more akin to the national income deflator, which incorporates all activities in the economy. Having agreed in principle to recommend the inclusion of services in WPI, it was felt that a number of practical problems would emerge if this were to be implemented immediately. Some of the problems that were anticipated are as follows:

1. The problem of availability of data on different services sectors on a continuous basis.
2. The problem of identifying services purchased by producers and services purchased by consumers.
3. The need for mixing and matching data collected from administered prices as well as from market prices for services rendered.
4. Problem of non-tradable services.

7.18 It is not possible, however, to correct for this major gap in the weighting diagram in one stroke as the services sector comprises many and varied types of activities. In the beginning, therefore, the Working Group recommends that an attempt be made to construct a separate index incorporating indices from the following sub-sectors:

1. Financial intermediation (including banking, insurance etc.)
2. Transport services (Road and Rail)
3. Communications (Postal and Telecommunication)
4. Water supply (Municipality rates for different groups; electricity, gas, etc. are already included)
5. Construction activity (in addition to the material inputs which go into them).

7.19 Except for financial services, data on the tariff structures in other types of activities listed above are available in principle. For the construction sector price data are expected to be available from various national, state and metropolitan level organisations such as HUDCO, DDA, HUDA, CMDA etc. Further work could be continued under the aegis of the proposed Standing Committee. As regards sub-sectors other than those listed above, the Working Group decided that the initiative to construct their price indices will be beyond the present agenda for revision.

7.20 Considering the above, it is recommended that in future (starting with the proposed standing committee) attempts may be made to develop a family of wholesale price indices. That may include:

1. the traditional commodities-only index comprising:

I. Primary

2. Fuel, Power light and Lubricants and
3. Manufacturing

1. Services (only those listed above) together with construction index as a separate entity.
2. A and B above combined to produce a composite index
3. A separate inflation related WPI measure by netting out intermediate transactions. This last variant (D) is discussed further below.

7.21 A problem that is endemic in any exercise of revision in the weighting diagram relates to changes in the quality of products and services. The Working Group has deliberated on this challenging issue at great length. It emerges that it is necessary to make a clear distinction between changes in the quality of a produce which do not materially alter the functional aspect of the product or service as distinct from one which provides, what can only be termed as, a new product or service. Thus, a minor improvement in quality satisfying essentially the same requirement in the process of economic activity or final consumption will be subsumed in the process of weighting the price by its value of production without any effect in the revision. However, a major change will be treated as a new product occupying an element of its own in the vector space of the basket of commodities and services. For instance, in the weighting diagram of the current series in use, even a product such as the Maruti car does not figure whereas in the version under revision it is expected to occupy a significant weight within the category of passenger vehicles. Many other competing products in the consumer durables segment have come in the market and they will also be reckoned as new products rather than merely satisfying the functional need of transporting passengers, washing clothes, clothing people, amusing children, etc.

7.22 It is quite true that an exercise to unscramble various dimensions of a composite product can be undertaken and a scheme of hedonic pricing corresponding to various components of the product can be identified. Developments in statistical and econometric techniques clearly show the way forward. For instance, a package tour may have several dimensions and changes in prices can be related to changes in the quality of various dimensions of that package. Similarly, a motor car, an aircraft or a designer outfit could be so priced. Although such an exercise could be carried out, and may also contribute significantly to micro-level studies undertaken for making behaviouristic inferences, it may not be a practical proposition for an exercise of the dimensions that the Working Group is contemplating. Moreover, the objective here is measurement and not the diagnosis of change or causal linkages. In any case, quality changes are a gradual and ongoing process for a majority of products and it will be foolhardy to attempt such an exercise for a weighting diagram which is expected to last for upto a decade. It is, therefore, considered prudent to make a binary choice between "no change" and a "new product" to tackle the problem of quality changes.

7.23 It is clear that the measure as it is followed so far is somewhat unsatisfactory because the WPI that is in use at present is intended to be an index of "all transactions", though it is only partly so, since "services" are excluded. Also the prices, as represented, are a mix of ‘producers prices’ and ‘wholesale market prices’. Conceptually a ‘measure of inflation’ should measure the prices only at the "final demand level" and not at "intermediate demand" level. If Iron ore gets into the production of steel, and steel gets into the production of machine, the price of machine and the changes in this price already subsume the prices of iron ore & steel and changes in those prices. In measuring changes in the price of all the transactions, we are counting the changes in price of ore, changes in price of steel and changes in price of machine. Thus there is a cascading effect. For true ‘measure’ of inflation we should measure the changes in the prices of ‘final goods’. The sum total of final goods also corresponds to the national income which is the sum total of final consumption plus trade. As far as the inflation measure is concerned, it is necessary that the proper weighting diagram ought to exclude inter-industry transactions that permeate economic activity but which are not central to the measurement of the price of final goods and services in the system. There is no point in including these if the objective is to measure the prices of the final vector of consumables. It is hoped that a meaningful index, as indicated in option (D) of paragraph 7.20 above, could be constructed as a separate measure of inflation. It is pertinent to note here that the proposed index in this context is not a substitute for the retail price index at the national level, although it may move in sympathy with that. It will, however, be close to the implicit GNP deflator.

7.24 Besides the several issues discussed earlier, there are certain other aspects which are quite important from the point of view of the index of wholesale prices. These issues, inter alia, include handling of seasonal items, classificatory system, method or the formula for calculating the index, linking of the old and new series, and the finalisation of index.

7.25 There are a number of agricultural commodities, especially, some fruits and vegetables, which are seasonal in their availability and whose prices are quoted only during a particular period of the year. Such seasonal item are handled in the index in a special manner. When a particular seasonal items disappears from the market and its prices cease to get quoted, the index for such an item ceases to be compiled and its weight is distributed over the remaining items within the concerned sub-group on a pro-rata basis. This system has been in practice in all the previous series and will be continued in the revised series also with a clear delineation of the specific period during which the index of a particular seasonal item will be compiled.

7.26 The system of classifying the selected items for the revised series engaged the due attention of the Working Group. After giving due consideration to the various aspects of the issue, the Working Group came to the conclusion that the classificatory system in vogue in the current series might be continued for the revised index also. The Working Group noted that if any change is made in the classificatory system, it might create difficulties in linking the revised series with the existing one. It should, in any case, be possible to rearrange the classificatory system to suit one’s requirements on the basis of the detailed individual commodity indices and their weights. Thus, the National Industrial Classification (NIC) being generally used in the existing series in respect of the major group ‘Manufactured Products’, has been retained for the revised series also. There are, however, modifications made in regard to the grouping and sub-grouping of the selected items for the purpose of taking into account the structural changes that the economy has undergone over time.

7.27 There is no change in the method of compilation of the index in the revised series. It is calculated on the principle of weighted arithmetic mean, according to the Lasperyre’s formula which has a fixed base-year weighting diagram operative through the entire life span of the series.

The formula used is:

Wi = The weight assigned to the ith item/sub-group/group/

major group.

Ii = Index of the ith item/sub-group/group/major group

7.28 Price relatives are calculated as the percentage ratios which the current prices bear to those prevailing in the base period, i.e., by dividing the current price by the base period price and multiplying the quotient by 100. The commodity index is arrived at as the simple arithmetic average of the price relatives of all the varieties included under that commodity. The indices for the sub-groups/groups/major groups/all commodities are, in turn, worked out as the weighted arithmetic mean of the indices of the items/sub-groups/groups/major groups falling under their respective heads.

7.29 The weekly index of wholesale prices at the time of its initial compilation and release is provisional in nature because it does not take into account some of the price quotations that are received belatedly. In such cases, the prices of the missing quotations are either repeated or estimated depending on the nature of the commodity. The provisional index is made ‘final’ after a period of eight weeks by which time almost all the required price quotations are expected to have become available.

7.30 The Working group decided that the present practice of first compiling the index on a provisional basis and then finalising it after eight weeks might continue in the revised series as well. It also suggested that in case some price returns are received late by more than eight weeks, then they would be used with effect from the week for which the index is being finalised in the current week and that any index earlier finalised should not be revised.

7.31 In order to maintain continuity in the time series data on wholesale price index, it is imperative to provide a linking factor so that the new series, when released, may be compared with the outgoing one. The linking of the two series is significant on two counts. Firstly, it provides a basis for the determination of cost escalation, wage settlement, etc. and secondly, it generates a long time series of data/information which can be used for undertaking certain analytical studies on the behaviour of prices.

7.32 There are several methods available for linking a new series with an old one. Some of the most common and widely used methods among these are: (1) arithmetic conversion method (2) ratio method and (3) regression method. In the arithmetic conversion method, the simple average of the monthly index numbers of the old series is expressed as a ratio of the average index for the same period (generally the base period) of the new series.

7.33 The ratio method involves taking an average of the ratios of the monthly indices of the old series to those of the new series for a given period.

7.34 In the regression method, the monthly index numbers of the existing series are regressed on those for the corresponding months of the revised series to give an estimated linking factor.

7.35 Out of the above three methods, the arithmetic conversion method is the simplest to adopt and often the most convenient method for operational reasons. The office of the Economic Adviser has been using this method to link various price index series. The choice of the method to use for linking various price index series is, however, left to the user.

7.36 The National Informatics Centre (NIC) and Working Group Division of the Office of Economic Adviser have been working in a cohesive manner for the successful release of the forthcoming Wholesale Price Index Series. The smooth transition from series 1981-82 to the new series 1993-94, has been mainly due to the sustained and devoted efforts of National Informatics Centre. The Techno-functional support from NIC has helped in the efficient implementation of the new series. For this purpose, many software modules were developed by NIC. It facilitated the Secretariat in focusing attention on the scrutiny of raw price data and its consistency check, generation of reports for decision making and the compilation of indices. The basic technical methodology for compilation of the WPI remained the same as for the current series. It is also proposed to make the revised series available on the WEB site of the Office of Economic Adviser, Ministry of Industry from the first release of the revised WPI series.