The non-correspondence issue as explained above was ignored at one stage for some exercises carried out by the working Group. Those exercises revealed a great deal of disjunction in weighting patterns at several important points between the two sets of methods (and measures). There were three major reasons for the Working Group to undertake a course correction and bring about a greater degree of conjunction between the two sets of measures at the level of the macro economy. First, the WPI weighting diagram is derived from a basket that contains only those commodities which are considered important to monitor the movement of price levels in the economy. In that they correspond well with the universe it seeks to represent. However in practical terms commodities which were otherwise considered important for inclusion in the basket had to be pruned due to non-availability of reliable data on a continuous basis with voluntary compliance. In view of this practical difficulty the possibility of the sample weights diverging significantly from the "true" weights at various levels of disaggregation was considered real. This possible source of distortion needed correction at least at some level of aggregation so that the sample (basket) designed is not forced by circumstances to remain out of tune with the universe it seeks to represent.

Secondly, the exercise of representing the value of output on its own platform of weights (i.e., aggregated from available data "from below") gave rise to certain outcomes which were not easily amenable to reasoning by the Working Group based as they are both on economic theory and historical experience. While it is true that a series of ‘repeated samples’ could in principle avoid this problem of divergence, such is not the practice of compiling the WPI weights, which are at pains to be reconstituted about once in a decade. Caution and prudence were therefore made to prevail over theoretical puritanism. A close correspondence of the weights at the aggregate level was thus sought by the Working Group to tune its reasoning with that of the fund of knowledge and wisdom available from the analysis of national income and related measures as higher levels of aggregation. Thirdly, the present Working Group has recommended many structurals breaks to be carried out in future. However not everything can be changed together in one fell swoop in order that discontinuity is avoided. Thus in continuation of the practice adopted by the last Working Group, correspondence of the weighting diagram with their counterparts in national income series was secured, in order that meaningful comparisons can be made with the structure prevalent at the time the 1981-82 sample basket was adopted.

It was decided that the shares of primary and secondary sectors as found in the GNP data with the new base year (1993-94) be adopted for the WPI weighting diagram as well, corrected for the non-inclusion of the services sector in the weighting diagram. Accordingly their shares, together with the share of Group II "Fuel, Power, Light and Lubricants", were imposed from outside the sample in the manner of parametric data. The shares of different sectors at the second digit level of NIC classification were also imposed from outside for similar reasons.

"Primary Articles" was thus sub-divided into "Food Articles ", "Non-food Articles" and "Minerals other than petroleum crude". The "Petroleum Crude" share was shifted to the second group which contains "Coal Mining", "Mineral Oils" and Electricity". However, as a conscious decision, petroleum crude as an explicit item has been dropped from the series, but its weight in terms of the value of trade has been included in the sub-division "Mineral Oils" and distributed among the list of included items on a proportionate basis; it is implicitly assumed thereby that while crude petroleum enters into all the petroleum products as determined largely by technological considerations, its weight cannot be ignored in view of the large value of trade that takes place in the economy.

The third group consisting of "Manufactured Products", which is sub-divided into its major component parts as they appear in the ASI classification (13 classes from A to M), and their weights have also been assigned from outside in the manner discussed above. For practical reasons the last class comprising "Other Miscellaneous Manufacturing" had to be dropped; the data available represented only ‘wrist watches’ and this item threatened to represent an entirely amorphous group with a combined residual weight of 3.366 per cent in 1993-94. The weight corresponding to this class was therefore distributed among all the other 12 classes pro-rata. To that extent there will be a slight deviation from the weights originally imposed from the national income data. However, their proportionality will be maintained by design. The deviations which would occur are rather insignificant by absolute measure, so that the structural characteristics are left unchanged.

The commentary that follows on the comparison between the current weighting diagram (1981-82) and the one proposed ( 1993-94) ought to be seen in the context of the methodology of compilation outlined above.

As the weights at the first two digit levels have been taken as they appear in National Income Statistics, mutatis mutandis, no comments are warranted for changes which occur at these levels. It may be noted that the adopted weights factor in the changes resulting from the excluded services sector, which the national accounts data include. Some comments at lower levels of disaggregation of item groups and individual items are however warranted.

As would be expected from the changing expenditure pattern of the econo- my the share of food grains has gone down in the revised index. This is quite prominent in cases of all cereals and pulses. However this trend would have been more pronounced if the marketed surplus ratios (MSR) had remained the same. From a comparison of the data available for the two sample periods, it emerges that the MSR have gone up significantly in virtually all cases in this group. Similarly, a comparison of the trends between Fruits and Vegetables together as a group and at item levels shows the same declining weights emerging in these commodities. This is however not counteracted by the changing MSR in these items; if anything, a number of items have been observed to have a declining MSR . Nor is this trend countered, it appears, by the introduction of new items in these groups. In the case of ‘milk’ however, the trend is very prominently reversed; there is more than doubling of the weight assigned to this product in the new sample as compared with its predecessor. This is only partly due to the increase in the observed MSR from 60 percent to 100 per cent between the two periods.

Within the category of non-food articles, all the groups, namely "fibers", "oilseeds" and "other non-food articles", have seen decline in their weights. The only exception to this trend have been "raw cotton" and "raw rubber", which have guarded their positions quite firmly.

In the case of "Minerals" within the "Primary Articles" sector, a significant change of practice has been effected in that the weight of petroleum crude has been shifted to the second sector, i.e., included among "Fuel, Power, Light and Lubricants", thereby effecting major changes in their respective shares. As a result of this change, the second sector has increased by about 4 percentage points, almost entirely attributable to this change of practice. Within the group of ‘mineral oils’, which as a group has maintained its position, LPG has increased prominently whereas other items have managed to cushion themselves with minor changes in their respective shares. Significantly, petrol and HSD have maintained their positions over this period of a dozen years. It may be recalled that some major changes have taken place in POL price regimes between the two periods. It is also seen that the share of electricity as a group has doubled over the period, although data available with the Working Group do not allow for detailed comparisons of the change.

Moving on to the "Manufactured Products" sector, some interesting changes have taken place in the dozen years. Among the highlights, it is seen that "maida" and "atta’ have changed their relative positions, with the former surprisingly dominating in the value of trade in companion with "atta’. No further comments need be made here since they reflect a similar pattern as per the source data (the ASI data). The share of "sugar" has increased very prominently in line with the changing clout of this product in the economy. Within the same category of "Food Products", other prominent changes are in the groups "Edible Oils", "Oil Cakes" and "Tea and Coffee processing", all of which have notched up increased shares.

In the category of "Beverages, Tobacco and Tobacco Products", the "Manufacture of Bidi, Cigarettes and Tobacco" group has recorded a prominent decline from nearly 2 per cent to nearly 1 percent.

Within the declining category of "Textiles", the prominent change has been the decline in the shares of "Cotton Textiles" and the increased shares of "man-made Textiles". "Woolen Textiles" and "Jute, Hemp and Mesta Textiles" have also made some losses in their shares.

The weight of "Wood and Wood Products" has declined prominently , while the group "Paper and Paper Products" has maintained its weight along with the group " Leather and Leather Products". As would be expected, greater weights are now attached to the products belonging to the group " Rubber and Plastic Products". "Plastic Products" have gone up, and among "Tyres and Tubes, "giant tyres" and motor tyres have increased significantly.

There has been a 4 percentage point increase in the category (imposed from outside) of " Chemical and Chemical Products", within which a 2 percent point increase has been effected by the products comprising " Fertilizers and Pesticides". Within the same group, "Drugs and Medicines" have also notched up prominent gains.

The only event worth noting in the category of "Non-Metallic Mineral Products" is the prominent increase in the weight of cement in the economy. While the category "Basic metals and Alloys" as a group has increses its share along with the group "Non-Ferrous Metals" which have been given way to by "Metal Products". Within "Basic Metals and Alloys", "Iron and Steel" and "Pipes, wires, drawing and others" have increased, while foundries for castings, forgings and structurals" has gone down. Among "Non-Ferrous Metals", aluminium has increased its share.

"Machinery and Machine Tools" as a category has increased (non-sample data), taking along with it the weight of electrical machinery, while non-electrical machinery as a group has increased, albeit marginally. Within the group of "Electrical Machinery", "Electrical Industrial Machinery", "Wires and Cables" and "Electrical Apparatus and Appliances" have increases their shares prominently, together with the introduction of new items and products within their ranks.

Among "Transport Equipment and Parts" as a category, which has increased prominently over the period, "Motor Vehicles, Motor cycles and Scooters" as a group has increased its weight. Within this group while "Truck chasis" and "Bus chasis" have increased, "Car chasis assembled", "Motor cycles" and Scooters" have increased prominently. The group " Locomotives , Railway Wagons and Parts" has also registered increases during the period. The group "Other miscellaneous manufacturing" has also increased its share quite prominently from 0.972 to 3.366 as a residual group, but this has been distributed over the other remaining groups due to the lack of detailed data within this amorphous group.