Wilfred Malenbaum Economic Development Program C/54-35 India's Domestic Product. 195L/52--15 Wilfred Nalenbaum In the course of studies on the process of economic growth, now being carried on at the Center for International Studies of the Massachusetts Institute of Technology, estimates were made of India's domestic output in the years since 1948/9. The procedures used were relatively crude; moreover, they could be applied only with such information as was available in Cambridge. The results were considered approximate, and are used as such by the Staff of the Center. They are presented here in the hope that examination and criticism of the method and statistics could lead to improvements in the estimates. The results might then warrant broader use for wdrking purposes, pending subsequent reports of the National Income Committee. The estimates of domestic product are shown in Table I, in current and constant prices. The procedure used yield a result at the market price level; further adlustments would be needed to convert the figures to a factor cost basis. Table I Preliminary Estimates of Indian Domestic Product (Market Level . n oun Prime 1952-53 91.6 96.8 100.3 101.1 97.0 1953-54 101.7 19j8-69 1949-50 1950-51 1.951-52 Re, abja) In 1484 91.6 94.4 92.1 87.5 95.8 96.A -Caloulations in current prices were deflated by the Economic Adviserts 2W10. General. Wholesae Price Indes, with 1948/U The entire series is shown, although official estimates for 1949/50 and 1950/51 have since become available from the National Income Unit. For the overlapping years, the present estimates are very close to the official figures of "net national product at matket price." However, the NIC's Final Report has revised the earlier estimatefor 1948/49, upon which the 1 lndireet taxes and miscellaneous fees average some 500 Res. are over these years. There remains the problem of allocation of these to various companies of the domestic product. 2 present calculations are heavily dependent. adjustment. 1 They will therefore need some Table II Preliminary NatUonal Product Estimates by Industrial Origin (arket prices, Rs. abja) Agricultur Industry Faetory (and mining) Small Enterprise Trade 1948/49 1949/50 190/51 42.8 46.7 48.7 48.6 43.2 7. 9.2 7.6 9.2 7.8 9.0 9.5 8.4 9.7. 8.9 18.1 18.8 19.4 18.5 19.1 16.1 161. 101.1 97.0 Services Domestic Product 91.6 96.8 1003 1951/52 31952/53 In Table II are shown the industrial origin breakdowns of the figures, in current prices at the market level. The breakdown was not carried further than the four (or five) sectors shown-for reasons which will be clear .in the discussion below. One can again observe the correspondence with the same sectoral breakdown, 1948/49;;-195/51, in the Final Report of the NIC (bearing in mind, again, that present figures are at the market, rather-than the factor cost, level)* Revision of the present estimates on the basis of the new information can be expected to yield an even greater correspondence by sectors, and, on the average at least, for domestic product as a whole. Derivation of Me stiate The procedure is basicAlly that of extending the NIC 1948/49 estimates through time, by the use of appropriate price and quantity indexes. The indexes are applied to gross output (or tales) in each sector, and to the intermediate costs paid by this sector to all sectors of the econmy for goods and services "used up' in the productive process. The procedure is described, briefly, in the folloitg sections. lIn the present situation, these adjustments are bound to be mall since the NIC figures referred to are, in Re. abja for 1948/49, 19.9/50, 1950/51 respectively, 90.6, 94.6, and 100.3. In general, however, such estimates as those presented-here should take off from the latest and most complete official calculations available. 1953/54 47.4 10.3 8.4 19.3 121.7 3 (a) The cost matrix in 1948/49 Table III Estimated Gross Output and Costs, L94&/49 (Market pries$ R. abja) A. Costs Incurred In: Indust Trade Faltory ma~l Mining Entreprises Services Total For goods and services from: 10.27 4.80 4.70 .72 .35 20.8 2.08 2.48 2.80 .3140 .70 9.A Trade .46 4.57 1.90 2.11 .28 9.3 Services .03 .28 .02 .20 .01 .5 Agriculture Industry Total 12.8 12.1 9.4 4. 1.3 40.0 B. Value Added* 42.8 7.4 9.2 18.1 14.1 91.6 C. 55.6 19.5 18.6 -2.5 15.4 131.6 Gross Output: Table III represents the results of an effort at reconstructing the gross output estimates provided in the Firstp the NIC. This nt the l A) aCated mong the four broad estimation of. intermediate coats (S-e sectors to which they were paid.~ Lniessence, the process is the reverse of what the national income estimator uses: he starts from gross output and sales, and attempts to reduce this to a not basis. This means that Table III could undoubtedly be constructed with much greater ease and precision from the information availabe in NIU files and working papers. The primary source of raw paterial was the First Report itself, and primarily the detailed Notes and Sources section ofit p Some guidance was also obtained from a similar table constructed by Mukherjee for 1949/50. While this was developed for different purposes, Mukherjee makes clear that he was corronted by the identical problems in obtaining intermediate cost estimates. ISee, the Technique of Social Acgounting in a Pro-Industrial Economr, a paper read at thW third conference 9 f the International Association for Research in 1Icom and Wealth in Castelgandolfo, Italy, September, 1953. The tabe an dscussion on its construction are on pp. 14-16 of the iuimeographed copy in Cambridge, Mass. Since the sources for his 1949/50 calcuations were not available at the Center, these could give little specifiq guidance in our estimates for 1948/49. However, Mukherjeeis general orders of magnitude were a helpful guide. In Table III, Section B shows the figures of net output by sectors from the First Repot. They were augmented by an approximate allocation of indiretaxes paid in that year, so that they might represent the market value of the value added in the sector. Section C is of course the sum of A and B. ith the exception of the breakdown of industry into "factory establishments" (plus mining) and "smal enterprises." the categories used are the main groupings of the NIC Rort Broader groupings of industrial origin on of intermediate coets. On the other sectors complicate the est groupings selected were also influenced by the purposes hand, for present the availability of price and quantity indexes by which the various entries in Table III- can be extended to later (or earlier) years. (b) Extention of gross output and intermediate costs over time. The objective here is to estimate a Table III for each of the years The extension of Section G over time presented the least problem, although here considerable arbitrariness was necessary, at least on the basis of the information available at the Center in Cambridge, Gross output in agriculture could be projected by the used of a value series which is the product of two familiar seriess . wholesale prices of agricultural commodities (issued by the Ministry of Commerce and Industry) and th index of agricultural production (prepared by the United Nations). For industry, the comparable step was based upon the familiar series of wholesale prices of industrial comnodities and the interim inder of industrial nroduction, both from the Ministry of Commerce and Industry. 1949/,0 thraugh 1953/56. For the two broad sectors that remain-"-comprising commrce, transport and commications; and government and other serives, respectively-there are, to our knowledge, no comparable series to serve as a basis of projection. In, 1948/49, trade and services accounted for 29% of the gross 1 1n all cases, the indexes were first converted to 1948/49 - 100. actually, tne procedure described above was modified in our work: agriolture was separated into crop and livestock sectors (not shown in the zanles); industry into its large and small components (as in Tables II and III). The above procedure was actually applied only to tne crop section of agriculture (about 65 per cent of agriculture's gross output in Table III) and to the large industry component of industry. For livestock and small-scale industry, separate quantity series were estimated, derived largely from livestock census data, and from bits and pieces of information on small-scale industrial output. This is an area where more investigation is obviously needed. In the case af agriculture, at any rate, the uwo mstnods (working witn agriculture as a whole, and with cropEs and livestolk separately). show relatively small disparities. output estimated in Table III--a relatively larGe part of the total. In the absence of current: series applicable to these seators, the altermatives were (a) to maintain the 2) per cent ratio for later years, on theb groundtnat these two sectors would tend to vary with Un total of agriculture and industry, or(b) to- venture sOme preliminary estimtes on the basis of miscellaneous series available for components of trade and services. The latter procedure was used,- 4tlugh here again the two procedures provide not very different results. Oy these rug& methods, a Table III, Section C was constructed for the years after 1948/h9. For intermediate. costs, the estimates in Seon A were extended on the basis of tbe very indezes described above. However# some were applied in a less automatic mannnr. For example, agrulture's purchases of intermediate goods from industry (estimated in Table III at a total of Rs. 2.08 abja) were projected through the application only of the industrial price index. The argument here was simply that most in dustrial input factors in agriculture tend to be sticky; farmers do not alter their purchases directly with variations in output of agrioulture or industry. On the other hand, the goods and services purohase by largesoUeIndustry from agriculture (estimated at Rs. 4.80 abja in 198/49) we&varied in accordance withindustrial gross output. The basis for this was the argument that raw materials used by industry would in fact tend to vary more or less with industrial volume, and would be influenced by prices of (final) industrial output. - IA most oases, however, inter-' mediate costs were varied in accordance with the indexes developed for the sector in which the costs were incurred. Thus, tradesa purchases from industry were assumed to vary as did the value of total industrial output. (a) The final 'computations Table 111A is an end-result of the above processes. It is the parallel of Table III for 1953/54, and is simply the last of those prepared for the years from 1948/49. Section B i.s, of course, obtained as a residual from This of course dos not at all imply that either or both give reasonably -valid results. If the present method of estimating domestic product is to be used, series for trade and servics .will be needed. For whatever interest they may have, the present estimates extended gross output over the years from 1948/49 by applying the following indexes to the gross outputs in Section C of Table IIIM Trade Servises 1%18/14W 195 3 *lo 107 103 109 115 117 104 195 307 109 114 6 the direct estimates made for Sections A and C, as indicated above. It yields all the figures, at the current price level, shown for 1953/54 in Tables I and II. Table IIIA Estimated Gross Output and Costs, 1953{54 arket prices, Has. aba) A. Cste incurred In; A industrY -Fai , mall Trade Servioes Total Mining Enterprise. Par goods and services from: Agriculture 10.66 6.60 2.27 3.40 Trade .46 6.30 Services *03 .0- Industry Total 131. 47,o4 10,3 C. 60.8 26.9 .80 -40 23.36 36A20 1.90 1.00 11.6? 2.60 2.30 .30 11.96 2 5.2 3.7 47.o5 8.4 19.3 16.3 101,7 18.0 149.2 10.6 B. Value Added: Gross Output: .90 19.0 24.5 Some Observations on the Results Since 1950/1, real domstic product has increased by almost 5 per cent (after a sharp reduction in 1951/52). This is essentially the sam rate of progress projected by the Planning Commission in the First Five. Year Pla (pp. 2023). This stands in some contrast to the f=41istate.mens: on shortfalls of investment over the past plan years. Is the difaference due to undrestintion of actual investment or to more favorable factors, like favorable Weather and the existence of underutilied capacity, and'how is the basic improvement in India's econoda potential following from the development efforts? (b), The pripensity to import If one compares the series of national product estimates in real terms (second column, Table I) with the official quantity index of Indian imports, one is struck with the fact that the relationship seems to be inverse. The highe4t quantum of imports was in 1950%51, and the smallest in 1953/54, for eah-ole. The statistics sew*t6 be at odds with the frequently stated proposition cir timttwepeUities in underdeveloped areas. There are of course only six observations in the present case, and the food import 7 situation introduces special considerations. Nonetheless, the sharply re- duced demand for grains from abroad in 1952/53 and 1953/54 did ease the foreign exchage position. Increased incomnes did not result in increased use of these resources. One must of course take account of the effects of import controls. Stin, even these few statistics point up the need for further examination of consumption functions in India. This has significance not only for Indias imports but for the possible effects of a rapid expansion in domestic industrial output. (a) The sectoral pattern of changes in output One might find in Table II some evidence of a reduced importance of the value of agricultural output in national product in years of good harvests. This was most marked in 1952/53' Increased physical output in agriculture tends to be offset by price declines. Domestic terms of trade may also turn against agriculture. Such statements, based upon .broad Components of national product, are countered by observations good crops in recent years are said to have improved the economic position of most agriculturalista. Whatever the true situation, the figures do point up the danger of applying national price variatiofts to the output of an agiculture in which a very large proportion of product is utilised by the producer or in his imediate area. One needs to work with sectors which are much narrower than agriculture, and which take into account both agricultural output and other forms of output in the rural household. However useful the beat estimates of total national product (and the present figures are certainly not these), they can throw only limited light upon basic economic relationships in many underdeveloped areas. Social accounts are needed-and over time-for sectors f the economy which are reasonably homogeneous from an economic viewpoint, lIn this connection, see also M. Mukherhee, p