(CATALOG NO. 11) This paper is prepared for staff use and is not for publication. The views expressed are those of the author and not necessarily those of the Bank. INTER1dATIONAL BANK FOR RECONSTRUCTION AND DEVELOPMENT Economics Department Working Paper No. 42 An Analysis of Capital Flows Between the Agricultcural and Non-agricultural Sectors of India June 6, 1969 This paper originated, as part of the work undertaken in the Fiscal Policies of Develop- ing Countries Division for a paper on "Capital Flows and Income Transfers Within and Between Nations to Sustain the Agricultural Revolution,," (Sec M69-290 dated June 2, 1969) whiich Mr. McNamara presented at the Bellagio Conference on Agricultural Development. The analysis is based on a methodology which was developed jointly with Mr. Per Eklund who applied it in a parallel study of capital flows in West Pakistan (Working Paper No. 41). I am indebted in particular to Mr. Wouter Tims for support and criticism in the preparation of this paper. An earlier draft version has received valuable suggestions from several colleagues in the Bank particularly Messrs. Ablasser, Myint, Kavalsky and WiJaide. Fiscal Policies of Developing Countries Division Prepared by: Emmerich M. Schebeck ANi ANALYSIS OF CAPITAL FLOWS BET1WEE THE AGRICULTURAL AND NON-AGICUJLTURAL SECTORS OF INDIA I. Introduction 1. The purpose of the study is to determine the direction and magnitude of capital flows between the agricultural and norn-agricultural sectors of India within the context of rising agricultural income., supp- lemented by subsidies, which almost entirely escapes direct taxation. Not only is direct taxation of agriculture extremely low, it is even declining with the abolition of direct taxation in several states, thus contributing indirectly to a further rise in income in the agri- cultural sector. 2. Information on the flow of funds between the agricultural and non-agricultural sectors of India is limited. FAO in its Indica- tive World Plan for 1962 to 1975 has concluded that although "the savings rate in agriculture proves to be lower than postulated for the entire economy, the agricultural sector could finance the industrial sector to a significant extent."l/ 3. The study, which in Part II attempts to analyze these capital flows shows that a large and rapidly increasing outflow of capital from India's agricultural sector appears to be taking place just when agri- culture requires a large volume of investible funds for rapid expansion of food grain production. Details of conclusions and policy implications relating to this capital outflow are discussed in Part III of the study. 1/ FAO, "Main Conclusion and Policy Implications of the Indicative World Plan Regional Study for Asi12" Ninth FAO Regional Conference for Asia and the Far East, Bangkok, Thailand, November 1968, p. 50. -2- II. Analysis Growth of Agriculture l>. Net output of agriculturel/ for the period 1960/61 to 1968/69 has been growing at an annual compound rate of 1.2 percent on a semi- logarithmic trend.Z/ During the first four years (1960/61-1963/64) of the Third Plan (196o/61-i96l/65) the growth of agricultural value added was 2.2 percent compounded annually on a trend basis. However, the severe drought of 1965/66, and the sharp decline of agricultural output depressed agricultural growth and resulted in an annual compound growth rate of 1.76 percent in the entire Third Plan period. Thus, the gap between projected production grovth of 3 percent for the Third Plan and actual achievements was substantially widened. Table 1: Growth of Agricultural Value Added in India l960_/6_1to6 1965/69_ (at constant 1964/65 factor prices) Trend rates of growth per annum2/ 1960/6l-196T765 1960/61-1965/66 1960/61-1968/69 - in percent - Major, minor crops and livestock 2.21 1.76 1.19 a/ Least squares estimate of "b" in the equation log Y = a + b.time. Souirce: Computed from Revised Estimates of National Product, Central Statistical Organization, GOI, 1969 unpublished, to be incor- porated in the 1969 White Paper. The new revision constitutes a considerable downward revision of previous estimates. 1/ Refers to value added of major, minor crops and livestock at constant 1964/65 factor prices. 2/ Because of the fundamental revision of the national product estimates by the Central Statistical Organization, which covers only the period 1960/61 to 1968/69, it was not possible to extend this analysis over previous years. -3- 5. Agricultural output during the Third Plan fell far short of the projected growth because of severe weather conditions and neglect of agricu'lture in the Third Plan in its reluctance to allocate to agriculture the resources necessary to accelerate adequately the growth of agricultural production. Particularly deterrent to growth was the inadequate supply of inputs, the absence of detailed planning, and the dispersal of limited resources over large areas .1/ 6. Changes in agricultural policy and programs, reflected by increased public expenditure in agricultuare cluring 1965/66, and the adcption of new technologies associated wi.th a nearly doublinlg of fertilizer consumption in 1966/67, and h-igher prices paid to farmers created conditions favorable for a potential rapid growth of the agri- cultural sector. A second severe drought in 1966/67 caused some setback to growth but in 1967/68 the value added in agriiculture increased sharply. The Central Statistical Office estimates a real growth (of value added) in agriculture by about 3 percent in 1968/69. 7. The Fourth Plan, which after 3 years of delay is supposed to be implemented beginning 1969/70, aims at an annual compound rate of growth of 4.6 percent in gross value added at factor cost. Although a growth of this magnitude is a substantial acceleration compared with the trend 1/ The Third Plan Progress Report, 1963/65, p. 47. GOI Planning Commission. value of 1.2 percent from 1960/61 to 1968/69, it is not an unrealistic proposition in light of the recent deve.Gpment and the rapid adoption of new techmology in the agricu-ltural sector. The main contributions to growth are expected to come from new irxrigation systems, increased appli- cation of fertilizer and adoption of high yielding crop varieties. A striking evidence of this development is the installation of 70,000 new tubewells during 1966 and 1967 which compares with only 80,000 during the Third Plan, thus adding about 1½ million ha to the net area irrigated.Y 8. Table 1 of the Annex shows the trend values from 1960/61 to 1968/69 of gross value added at facto.)r cost in consta-nt 19l/65 prices. Based on the trend value for 1968/69 of agricultural value added of Rs. 10,,128 crores at 1964/65 prices and the projected growth rate of 4.6 per- cent for the Fourth Plan we estimate a value added of Rs. 12,682 crores for 1973/74 and at Rs. 13,265 for 197'4/75, assuming a continuation of the projected growth rate. Estimates of Investment Requirements for Ag ulture 1964/65 to 197/7 9. Because we can assume that a potential for an accelerated growth process in the agricultural sector was created at the end of the Third Plan and investment decisions were geared to this objective, we choose the 1965/66 trend value of value added as the benchmark. The potential growth of agriculture from 1965/66 to 1974/75 is estimated by fitting an exponen- tial trend from the benchmark value of 1965/66 to the projected level of value added of Rs. 13,265 crores in 197L/75. (See Annex, Graph 1). A potential growth rate of 3. 45 percent per annum was fouund, and the possi- bility that investment decisions were made to attain this growth seems very likely. 1/ FAO, "Hain ConcluLsion and Policy Implications of t'ae Indicative W4orld anthe Far t Bangkok, ailad,Novemobna Co196 erence f8, p1a0a an.d t.he Far Ea..'3t, Ba-lkc.kg r'ailand, November 1968, p.10. 10. Because this growth rate of 3.45 percent will .not only be induced by investment inputs but also by current inputs, any investment estimate based upon this growth rate and a marginal capital outp-ut coefficient will result in an upward bias. The impact of current inputs upon agricultural growth must, therefore, be estimated to determine a rate of growth net of the impact of current inputs which can be entered into equation (1). From production coefficierits, and actual data and projections of fertilizer con- sumption, adoption of improved seeds and plant protection, we derive at a rough estimate of additional value added arising from the incremental use of current inputs. On this basis we estimate that from 1965/66 to 1974/75, about l.4 percent of the annual growth wvill result from current inputs.. Stated differently, about 40 percent of. the annual grbwth during this period is projected to be accounted fLr by current inputs. Hence, of the.potential growthi of 3.45 percent per-annum about 2 percent can be expected to result from fixed investment inputs.. 11. Investment requirements for agriculture are estimated from gioŽwth in value added and a marginal capital output coefficient. Total investment in agriculture required to produce a fixed output can be written as: IA = A.Vt (g-l)g (1) in which (IA) represents total required investment in agriculture, (C) the marginal capital coefficient, (Vt:) the gross value added net of the impact of current inputs, (g) the growth index of the gross value added in agricul- ture net of the impact of current inputs and (g-l) the growth rate.1/ 1/ Wouter Tims, ntical Techniques for Development Planning, A Case Study of Pakistan's Third Five Year Plan. Pakistan Institute of Development Economics, 1968, p. 92. -6- Total investment in agriculture (IA) is composed of public invest- ment (Ip) and private investment of the agricultural sector (Ipr) hence, IA = Ip + 'Pr 1" (2) jKginal Capital Coefficients 12. The danger of using capital coefficients to estimate investment requirements from growth in value added are well recognized, but we adopt this methodology in the absence of a more accurate estimation. Beside capital coefficients and growth in value added, an additional variable deter- mining investment is the time lag between investment and output. An increase in the time lag will increase the investment requirements. If investment expenditures in year to are equally distributed over the year, and it takes on the average 1½ year before the investment generates output, a time lag of 1 year between the investment year to and the full output year t, is implicit. In our analysis we assume that the time lag between the investment year and the full ou?tput year is at the average 1½, years in the agricultural sector. Hence, investment in year to is related to a production increase between t1 and t2.2/ / (Ip) consists of public domestic investment (Ipd) and. investment ex- ternally financed (F?), P= IPd + (3) No data of foreign aid disbursement to the agricultural sector are availa- ble for India. We have to neglect, therefore, equation (3) and consider only public investment (Ip). 2/ Wouter Tims, Analytical Techni s for Development Planning, A Case Study of Pakistan's Third Five Year Plan. Pakistan Institute of Development Economics, 1968, p. 88. -7- 13. The marginal capital coefficients for the period 1961/62 to 196h/65 are estimated from trend values of agricultural outp-ut and investment.l/ This period was chosen, because an inclusion of 1965/66 and 1966/67 into the trend calculations would have resulted in large deviations of actual output observations which were entirely the result of droughts and not related to investment activities. 14. Public investment expenditures by Central and State Governments and Union Territories for major, minor crops and livestock were estimated from actual investment data as reported in the Third Plan Progress Report of the Planning Commission.2/ Information on private sector investment is very fragmentary for the period under concern. Data of private fixed capital formation by cultivators and non-cultivators were only available for 1961/62. Based on the All-India Rural Debt and Investment Survey 3/ they are estimated at Rs. 166.8 crores in 1961/62 and constitute 48 percent of total investment. For the years 1962/63 to 1964/65 we assume that the share of private invest- ment remained constant. Table 2 shows public and private investment figures and trend values of total investment based on a semilogarithmic trend: 1/ The marginal capital coefficients are estimated from 1961/62, which is the first year of the Third Plan and the only year f'or which an estimate of private fixed capital formation is available (see paragraph 14). 2/ They are exclusive of current expenditures and include investment for agricultural programs, major and minor irrigation systems, flood control, rural electrification and rural wTorks. 3/ "All India Rural Debt and Investment Survey, 1961/62,," Reserve Bank of India Bulletin, June 1965. -8- Table 2: Public and Private Investment in Agriculture l1/62 to l&~ Rs. in Crores 1961/62 1962/63 1963/64 1964/65 - in current prices - Public Investment 180.2 197.1 226.2 283.0 Private Investment 166.8 181.9 208.8 261.0 Total Investment 347.o 379.0 435,0 544.0 - in constant 1964/65 prices - a/ Public Investment 203 212 236 283 Private Investment 187 196 217 261 Total Investment 390 408 453 544 Trend Values of To- tal Investment 377 421 470 525 a/ Index of invest- ment cost at 1964/65 prices, (Planning Com- mission). 89 93 96 100 Source: The Third Plan Progr Report, 1963/6o, GOI Planning Commission. "All India Rural Debt and Investment Survey, 1961/62," Reserve Bank of India Bulletin, June 1965. 15. Table 3 shows agricultural output (gross value added) in constant 1964/65 prices and trend veaues net of the impact of current inputs. From 1960/61 to 1964/65 output shows a growth rate of 2.2 percent per year on the trend value. Again on a trend basis, it is estimated that from 1961/62 cur- rent inputs accounted for about 20 percent of the estimated annual growth of output. -9- Table 3: -gricultural OuLtpu gss value added) ~ in Crores 1960/61 1961/62 1962/63 1963/64 1964/65 1965/66 1966/67 Agricultural Out- put 9,313 9,527 9,329 9,510 10,396 Trend Values of Output 9,197 9,400 9,607 9,819 10,036 (10,258) (10,485) Trend Values of Output net of .i.rnpact of cur- rent inputs 9,400 9,566 9,736 9,908 (10,083) (10,261) lNoUe: The figures in brackets are extensions of the trend. 16. Based on the trend values shown in table 2 and 3, and the assump- tion of a 1½2 year time lag between the investment year and the full output year we derive the followiing marginal. capital coefficients for agriculture: Table 4: Marginal Capital Coefficients in Assumed Total % Rate Marginal Caital Coefficient time-lag investment of output Total Average Average in years (Rs. crores) growth average 1961/62-1962/63 1963/64-196)4/65 1.5 1,793 2.21 2.58 2.33 2.82 17. We will use in our model the four year average of the capital co- efficients of 2.6, which seems to be a better representation of the capital output relations than the capital coefficient of 2.8 at the end of the Third Plan period. In comparison, FAO 1/ estimates for eight Asian countries 2/ 1/ FAQ, ibid., p. 50. 2/ The FAO study covers Ceylon, India, Pakistan, Philippines, Thailand, Malaysia, South Korea and Taiwan. - lo - incremental capital output ratios of 2.6 for the period 1962 to 1975. Public Investment-Exenditures 1961/62 to 1973/74 18. Table 2 of the Annex shows public investment expenditures for major, minor crops and livestock from 1961/62 to 1968/69 as reported in the Third Plan Progress Report and the Annual Plans. These data are actual expenditures with the exception of 1968/69 which are estimates. The projected public investment from 1969/70 to 1973/74 is derived from the revised Fourtth Five Year Plan document and mnount to Rs. 2,137 crores at constant 1964/65 prices. This amounts to an increase of about 55 percent of public investment during the Fourth Plan as compared with Rs. 1,376 crores based on trend values over the past five years (1964/65 to 1968/69). On the trend basis, public investment shows an annual increase of 4.45 percent from from 1961/62 to 1968/69. Using the 1968/69 trend value of Rs. 300 crores at constant (1064/65) prices, public investment in agriculture during the Fourth Plan would have to increase yearly by about 12 percent to reach the anticipated target. Considering the financial problems which the government faces to mobilize the necessary resources to carry out its development program, it remains uncertain whether such a step-up of public investment in agricul- ture is obtainable. 19. Two assumptions for public investment during the Fourth Plan period will be used in the model: (i) a low assumption based on an extension of the 1961/62 to 1968/69 trend, thus assuming an annual increase of 4.45 percent; and (ii) a high assumption based on the Fourth Plan investment target, thus implying an increase of 12 percent per year from the 1968/69 trend value (see Annex table 2). For the analysis of capital flows in 1974/75 we assume a coiitinuation of the trend as outlined under the low and high assumptions. Savings of the Agricultural Sector 20. Monetized and non-monetized savings by individuals is estimated to have varied from 1963/64 to 1967/68 between a low of 6.4 percent to a high of 7.8 percent of NNP. Table 5: Monetized and Non-Monetized S2vings by Individuals 1963/6)4 i964/6s 1965/66 1966/67 1967/68 - in percent of NNP - Individual Financial Savings 3.7 3.4 4.l 3.2 3.0 Individual Physical Savings 2.9 3.8 3.7 3.2 3.6 Total 6.6 7.2 7.8 6.4 6.6 Source: IBRD, Economic Situation and Prospects of India, April 1969. TUnfortunately, no estimate exists on sectoral saving rates and, in parti- cular, the picture on savings of the agricultural sector is incomplete. 21. The rate of monetized saving by rural households has been estimated at no more than 2.,:5 percent of income on an average over 1950/51 to 1962/63, with a marginal propensity to save of 1.56 percent.l/ Because non-monetary savings are not included in these estimates, the rates of rural savings are considerably depressed. This is particularly true if one considers the pro- gress of agricultural development in India between 1950/51 and 1962/65. The above estimates of rural saving are also depressed by the fact that savings 1/ Uma Datta Roy Choudhury, "Income, Consumption and Saving in Urban and Rural India," The Review of Income and Wealth, March 1968. - 12 - in the form of buLllion and other forms of hoarding are excluded from the concept of financial savi-ngs of agricultural households.l/ 22. To obtain an estimate of the magnitude of total savings by the agri- cultural sector, we estimate savings for 1963/64 as a function of the average propensity to save (s) of the previous years income (Vt-l), of direct taxes (T) and of the previous years indirect taxes (T't_1) on the consumption of farm households. Savings (SA) can now be written as: SA= s(V t1-T-Tt ) (4) Under the low assumption we assume an average propensity to save of 5 percent and under the high assumption of 8 percent. For the following years, we esti- mate marginal propensities to save (s'). We define the marginal propenisity to save as: St - e( V - p + S2/ e t-2 2 where (e) is elasticity of the rate of savings with respect to the per capita income growth, (p) is the growth rate of the agricultural population, and (s) is the average propensity to save. The elasticity of the rate of savings was estimated from various country data at 4.00. The growth rates for India's agricultural population from 1962 to l975 is estimated by FAO at 1.8 percent compounded per annum../ Again, under the low assumption we assume an average propensity to save of 5 percent and of 8 percent under the high assumption and derive the following marginal propensities: 1/ Arun Ghosh, "Consumption and Savings," Economic and Political Weekly, October 19, 1968, p. 1,619. 2/ Wouter Tims, Indonesia Debt Simulation Models, IBRD/IMDF, unpublished. 3/ FAO, Indicative World Plan for Agricultural Deve vpen, Asia and Far East, Volume II, p. 2, Rome 1968. - 13 - Table 6: Estimated Marginal Propensities to Save of the Agricultural Sector Average Propensity Marg±nal Propensity to Growqth rate of to save save agricultural s s' population Vt1 Year p Vt-2 Low High Low Hi - in percJEnt- 1963/64-1969/70 1.8 1.19 a/ 5 8 2.6 5.6 1970/71-1974/75 1.8 4.6 b/ 5 8 16.0 18.0 a/ Based on the trend rates of grorth of 1.19 percent per year from 1960/61 to 1968/69. b/ Based on the projected growth of 4.6 percent during the Fourth Plan and the assumption of a ccntinuation of this growth during 1974/75. Hence, the achievement of the anticipated growth target of agriculture during the Fourth Plan would bring forth a substantial increase of savings. In com- parison to the above estimates, the Planning Commission estimates the marginal propensities to save for the economy as a whole at about 21 percent during the Fourth Plan. 23. The incremental saving can now be estimated as a St = s'(AVt-AT-AT t-1) () Because value added at factor costs excludes indirect taxes on factor inputs but ir-cludes subsidies on these inputs, we derive income net of taxes but gross of subsidies and deduct direct taxes and indirect taxes on the consump- tion of the farm household. Sale and excise taxes on agricultural produce are not deducted. We accept the widely held notion that the incidence of these taxes falls upon consumers of the non-agricultural sector and not upon the - 14 - producers. 1/ 24. The direct taxes (T) affecting Indian farmers are primarily the land revenue tax and the agricultural income tax. Land revenue declined in real terms from about 1X4 percent of gross value added in 1963/64 to merely 0.8 percent in 1966/67, whereas the income tax remained over nine years at 0.1 percent of gross value added (see Annex, Table 3). While, public investrzent. in agriculture increased on a tt'ena basis by )4.5 percent comrrpoundec` ?nnuall-.y durino t1fie 1961'Gs direct-tay rece±cptE declnaed'.by- 5.6 percent per year. This fact, as well as the large inequity of the tax burden between the agricultural and non-agricultural sector, is widely recognized by Central Government authorities. Because the likelihood of any increase of direct taxes seems very low, we assume in our projections that the trend value of direct taxes (T) for 1966/67, which is estimated at 1,000 million Rs., will remain constant over the projected period.2/ 25. The indirect taxes upon the consumption of farm households (T'), i.e., gasoline, kerosine, clothing, etc., are estimated to be 10 percent of all indirect taxes upon agriculture. Indirect taxes upon agriculture in their turn constituted fiom 1960/61 to 1966/67 on the average about 29.8 percent of total indirect taxes levied by Central and State Governments. Receipts from these taxes increased progressively and are estimated at 6.2 percent of agricultural value added in 1966/67 compared to 3.5 percent in V.P. Gandhi, Tax Burderi oml nAgricuure, lfar;aVrd Law School, 1966. 2/ For a detailed discussion of agriculture taxation see: F. Abbate, "'Agricultural Taxation in India," IBRD, 1968, unpublished draft. V.P. Gandhi, Taxes, Subsidies and Agricultural Production, Indian Insti- tute of Management, Ahmedabad, 1965. - 15 - 1960/61. On a trend basis, indirect taxes upon the consumption of farm households increased annually by about 10 percent. With a high income elasticity of consumption of rural households, we assume for the purpose of our projections a continuation of the trend up to 1974/75. In eGap of the Aricultural Sector 26. After having quantified the parameters of our model, we can write the investment-savings gap of the agricultural sector (K) as K =IA - Ip - SA (6) which can be revritten to K =X.Vtt(g-1)g - Ip -sA From equation (6). it becomes evident that if private capital formation in the agricultural sector (IA - Ip) is smaller than savings out of agriculture (SA), there will be a net outflow of funds to the non-agricultural sector. Or if (IA - Ip) is larger than SA, an additional inflow of public funds will be needed to meet the investment requirements necessary to achieve a certain anticipated growth target. In case capital transfers from the non-agricultural sector (SF) occur, thus increasing the funds available for agricultural invest- ment, the outflow will increase or the necessary inflow of funds can be reduced' by the amount of (SF). Equation (6) can be e)ftended to A P A SF (8) K = A - Ip - sA- s Because no information is available on the magnitude of capital transfers from the non-agricultural sector, we will focus our analysis only upon equation (7). Table 7: Capital OutLlow from the Agricultural Sector in India Rs. in crores at 1964/65 prices Investment Public Investment in Requirements in Agriculture Saving Capital Outflow 1/ Agriculture Ip Direct Taxes Indirect Taxes SA K =- - Year IA Low High T T' Low High Low Hg 1964/65 525 251 251 113 472 448 719 -174 -445 1970/71 593 328 376 100 837 521 809 -256 -592 1973/74 629 374 529 100 1113 725 1039 -470 -939 1974/75 645 391 592 100 1225 798 1121 -544h -1068 1/ Negative figures indicate positive outflow. - 16 - 27. The degree of uncertainty attached to the estimates of the para- meters would make it unrealistic to give anything else than a range for the net outflow of private savings from the agricultural sector. The range of this outflow under the assumptions of low public investment and savings and high public investment and savings is presented in Table 7. For 1964/65 the capital outflow from the agricultural sector is estimated to have been in the ranrge of Rs. 175 to 445 crores. In the subsequent years the outflow is estimated to increase substantially and will range from Rs. 255 to 590 crores in 1970/71 and Rs. 540 to 1,070 crores in 19714/75. 28. The magnitude of this outflow is substantial in relation to both the investment requirements and to the estimated investment by the public sector. Even under the low savings assumption, the domestic resource availa- bility from the agricultural sector will about equal the investment require- ments in this sector by 1970/71. By 1973/74, at the end of the Fourth Plan period, savings will surpass the investment requirements by 15 to 60 percent and by 20 to 75 percent in 1974/75. 29. In order to fully evaluate the policy implications which can be drawfn from this analysis, it must be emphasized that the net outflow is rather insensitive to plausible changes in the basic parameters of the model. Even a lower growth rate would not change the direction of the capita3 flow. because the reduction of investment requirements would be larger than the re- duction of savings. In addition, any increase in the contribution of current inputs to the growth of gross value added would reduce the investment requiro- ments and would consequently increase the outflow. This outflow wculd also be increased by any decline in the assumed time lag of 1½5 years between invest- ment and output because it would cause a substantial reduction of investment - 17 - requirements. 30. Further evidence that an outflow of capital from the agricultural sector is not implausible is provided by an analysis of branch expansions by commercial banks into rural and semi-rural areas in India. Over the past years commercial banks were able to substantially increase- their de- posits in rural areas. Estimates of credit deposit ratios indicate that funds are transferred from rural to urban centers. The credit deposit ratio in urban centers is estimated at 73 percent compared to 43 percent in riral areas. eaice, a large part of rural deposits with the commercial banks are not used for credit expansion in rural areas but are diverted to industrial centers.l/ 1/ "Debits to Deposit Accounts with Scheduled Commercial Banks 1961-65," Reserve Bank of India Bulletin, Vol. XXIII, February 1969, p. 141-169. - 18 - III. Conclusion and Policy Implications 31. With the anticipated increase in expenditures during the Fourth Plan the task of resource mobilization becomes of vital importance to eco- nomic development. The problem of financing Plan outlays makes a larger participation by the State Governments necessary. The States, however, are not willing to raise more resources either through increased taxationl/ or other means and find it therefore imnpossible to contribute a much higher share to the Plan than in the past. The financial constraint is aggravated even further by the Plan objective to reduce the net foreign assistance by half as compared with the level at the end of the Third Plan. This vicious cycle of financial constraints can certainly not be broken entirely by re- sorting to more deficit financing. Under these circumstances the overriding aim of the Fourth Plan -- growth with stability -- becomes a meaningless slogan, unless adequate dcmestic savings can be mobilized. 32. What policy implications can be drawn from our analysis in the light of the financial cnnstraints of the Fourth Plan? In an effort not to sacrifice the Plants priority of agricultural development and its impact on price stability, State Governments have to overcome resistance to fully utilize the taxable capacity of the agricultural sector to enable them to 1/ During the past years, land revenue taxces wyere abolished in many stc.tes, thus reducing the already low incidence of this tax even further, despite increased pruductivity of agricultural land'. The receipts from this tax. could have been mo±re elastic if there hed been assessments at short inter- vp.ls, thus taking account of changes in the price level as rloell as in pro- ductivity of land. Since the 193C'ssauch reassessments were constantly post- poned by State Governments. Hence, the land revenue has lost its effective- ness as a source of taxation of the agricultural sector. Similarly to the land revenue, the agricultural income tax is only levied in some states of India. This practice results in an extremely low incidence per hectare which is almost nil on the nonplantation agriculture. In the last months the Union Cabinet debated proposals to alter the direct tax structure and to introduce a farm wealth tax. These proposals were objected to in ths Parliament and among State Governments out of the fear than an increased tax burden for the agricultural ,sector might impair agricultural productiv- i b*b. - 19 carry out the plan outlays of public investment programs in agriculture. It must be realized that the large taxable capacity of the agricultural sector is hardly utilized at all, and an increase in direct taxation does not have a disincentive effect upon production as it is often argued by politicians, but rather an incentive effect. Furthermore, it should be empha- sized that part of the financial problem lies in the direction of controlling the level and the growth rate of subsidies to agriculture. As it is with any general subsidy scheme so is the case with India: the larger farmers are the main beneficiaries. But it is this group of farmers with higher incomes who account for a large part of the capital outflow. A complete elimination of the subsidy programs would have a disincentive effect upon product,ion. But subsidies on a selective basis benefiting only certain groups of small farmers or certain disadvantaged regions would restore most of these incentives. In 196L4/65, when direct taxes of the agricultural sector were about Rs. 113 crores, total subsidies paid to this sector were estimated at Rs. 50 crores, of which the subsidy on irrigation amounted to about Rs. 37 crores. Some of the subsidy programs might have been justified to provide the initial necessary incentive to farmers to adopt a new technology. As farmers become more aware of the high returns from new technologies, it is not conceivable that an elimination of subsidies will effect the further adoption of these technologies. Thus, the disincentive effect upon produc- tion can be regarded as minor.1/ Although measures to increase taxation and 1/ The recent elimination of feftilizer subsidies was a step in this direc- tion, but it must be recognized that this subsidy was rather small and the total subsidies granted to agriculture are still very large. - 20 - reduce subsidies are necessary from the point of fiscal policy, it seems unlikely that actions on this line will be sufficient to deal with the imme- diate need for raising funds for agricultural development. 33. It therefore becomes necessary for the Indian authorities to resort to means of mobilizing funds othier than by a sole reliance on fiscal policy.l/ The most effective way to reduce the capital outflow would be by voluntary mobilization of these funds through the savings institutions and the back-plowing of these funds into agriculture. A large outflow of capital from this sector to the rest of the economy is certainly not desirable at a time when agricultural development programs by the public sector are threatened by financial restraints and when agricultural credit institutions are faced witl a shortage of funds to meet the credit requirements. The dichotomy be- tween agricultural development and its financing becomes apparent if we con- sider that the capital outflow from agriculture is prima.rily the result of deposits by larger farmers, who not only are the main beneficiaries of sub- sidy programs and tax concessions, but are also the major recipients of credit from the agricultural credit insitutions.2/ 1/ An examination of these non-fiscal alternatives ha6 been presented in: IBRD, "India: Agricultural Credit Mission,'t Draft Report, March 26, 1969, pp. 62-77. IBRD, "Economic Situation and Prospects of India," April 18, 1969, p. 5o. 2/ In 1961/62, 13.3 percent of rural households with assets of more than Rs. 10,000 received 53.3 percent of total cooperative credit, whereas 53.0 percent of rural households with assets of less than Rs. 2,500 received only 11.7 percent of credit disbursed by cooperative institutions. - 21 - 3h. The Fourth Five Year P1lan, and the report of the "Banking Group for Formulation of the Fourth Five Year Plan on Cooperation," stress the import&nce of resource mobilization and the increase of deposits with the cocperative credit system, but provide no suggestions as to how this can be achieved. At present, the agricultural financial institutions in India (the cooperative credit system and the land development banks) are predominantly channels of budgetary and semi-budgetary funds as well as compulsory funds of captive lenders (i.e., Life Insurance Corporation, commercial banks).l/ 35. As long as these institutions consider that budgetary and semi-budge- tary funds are limitless and can be obtained at negative rates of interest, they are not inclined to show a dynamnic effort to mobilize their olwm resources. In addition, a policy of low interest rates to agricultural borrowers com- bined with other concessionary features makes it too expensive for these in- stitutions to raise a large part of their lcanable funds through voluntary mobilization of deposits. Hence, the most effective way to mobilize agri- cultural savings would be through increaseddeposit and interest rates. The expansion of branches into rural areas is another important factor and account- ed primarily for the increased resource mobilization by commercial banks. However, the high cost associated with the opening of new branches in rural arweas prevents the commercial banks from establishing an intensive branch coverage. It usually takes three to four years until a newly-opened branch becomes profitable. In contrast, agricultural credit insitutions already have l! Over a three year period (1963/64 to 1965/66), funds from the Government and Reserve Bank accounted on the average for 53 percent of total outstand- ing loans by the cooperative credit institutions. The dependence of land mortgage banks on funds from the Government, Reserve Bank and captive lende.rs was even higher and amounted on the average to 65 percent of total produc- tive loans. (See IBRD, "India: Agricultural Credit Mission," Draft Report, March 26, 1969, Annex 15 and 16.) - 22 - a large network of branches in rural areas and would, therefore, be in a more favorable position to mobilize agricultural savings. However, these institu- tions were never forced to rely upon their own resources, and consequently did not find it necessary to market such banking services as provided by commercial banks. They therefore lack the necessary confidence in them to compete effectively with commercial banks in mobilizing private savings. 36. Agriciiltural credit institutions can only improve their image if they are made more responsible for generating their own funds and if they show a more responsible attitude in their lending policies. As long as suchl changes, which certainly would be revolutionary for India's agricultural credit institutions, are delayed, the only way these institutions could compete with commercial banks in mobilizing deposits and thus reduce the capital out- flow from agriculture, would be by raising their deposit rates above those of tlhe commercial banks. Table 1 Gross Value Added in Agriculture a, Factor Cost - Rs. in crores at constant 1964/65 prices - 1960/61 1963/64 1964/65 1965/66 1966/67 1967/68 1968/69 1969/70 1970/71 1971/72 1972/73 197344 1974/75 1. Gross Value Added in Agriculture Y 9,313 9,510 10,396 8,913 8,912 10,463 10,777 2. Trend Values of Gross Value Added in Agriculture 9,215 9,547 9,661 9,776 9,892 10,009 10,128 (i) Projected annual growth of 4.6 percent (1968/69-1974/75) 10,594 11,081 11,591 12,124 12,682 13,265 3 Potential Growth of Gross Value Added fron the Benchmark of 3965/66 9,661 9,776 10, .13 10,462 10,823 11,297 11,583 11,982 12,396 12,823 13,265 gB 4. Potentia3. Growth of Gross Value Added Net of the lIpaot of Current Inputs 9,661 9,741 9,943 10,148 10,359 10,573 10,792 11,015 11,243 11,476 11,714 Revised Estimates of National Product, Central Statistical Organization, GOI, 1969 unpublished, to be incorporated in the 1969 White Paper. The new revision constitutes a considerable downward revision of previous estimates. INDIA:GROSS VALUE ADDED IN AGRICULTURE AT FACTOR COST AT CONSTANT 1964/65 PRICES (RUPEES IN CRORES) 15 1 5 1 4 1 4 1 3 .: 1 3 POTENTIAL GROWTH - -1 FROM THE BENCHMARK OF 1.96 .-l ACTUAL OBSERVATION Tm Yc (9215) (1. 19) 10 9 9 8 8 SEMI-LOGARITHMIC SCALE 7 I I I I I I , . ,-I. 7 1960i61 1961/62 1962/63 1963/64 1964/65 1965/66 1966/67 1967/68 1968/69 1969/70 1970/71 1971/72 1972/73 1973/74 1974/75 - - IBRD-4403 I Table 2 Actuals and Projections of Public Investment in Indian Agriculture Rs. in crores 1961/62 1962/63 1963/64 1964/65 1965/66 1966/67 1967/68 1968/69 1969/70 1970/71 1972172 1972/73 1973/74 197175 Public Investment (in current prices)2! 180.3 197.1 226.2 283 348 366' 373 364 Public Investment (in constant 1964/65 prices)!/ 203 212 236 283 319 295 268 262 Trend of Public Investment (l96L/62-l968/69)g' 219.8 229.8 240.3 251.3 262.7 274.6 287.2 300.3 (i) Righ Assumption (1969/70-1974/75)W 336 376 421 472 529 592 (ii) Low Assumption (1969/70-1974/75)-/ 314 328 3143 358 374 391 2I Exclussive of public expenditures for forestry, fishery, community development and cooperatiOa, but inclusive of investment expenditures for agricultural programs, da3or. and isinor irrigation, flood control, rural electrification and rural works. The net investment for agricultural programs is estimated from the Plan documents at 64 peroent of total public outlay for agricultural programs. Data from 1961/62 to 1969/68 are based on actual investment and for 1968/69 on Plan estimates. 2/ Deflated by the Index of Investment cost at 1964/65 prices. Source: Planning Co.mmission. 3/ Least square estimate of equation log Y a + b.time. Estimat. Ef 'b"' 1.0445. 4/ The high assumption is based on projected public investment expenditures during the Fourth Plan of Rs. 2,650 crores at 1966/67 prices or Rs. 2,137 crores at 19614/65 prioes. 5/ The low assumption is based on an extension of the trend (1961/62 to 1968/69) thus assuming an annual increase of 4.45 percent and a total public investment of Rs. 1,717 aroma at 1964/65 prices during the Fourth Plan. Table 3: Direct Taxes on Indian Agriculture and Indirect Taxes on the Consumption of Farm Households 1960/61 1961/62 1962/63 1963/64 1964/65 1965/66 1966/67 - Rs. in crores at current prices - Land Revenue 97.78 95.79 120.65 123.70 119.85 112.05 88.05 Agricultural Income Tax 9.49 9.4h 9.59 9.26 10.73 10.24 10.57 Total Direct Taxes 107.27 105.23 130.24 132.96 130.58 122.29 98.62 Indirect Taxes on Consumption of Farm Households 28,214 29.88 36.18 44.69 44he74 60.00 68.00 - Rs. in crores at constant 1964/65 prices l/ - Land Revenue 113.70 107.63 129.73 128.85 119.85 102.80 71.01 Agricultural Income Tax 11.03 10.61 10.31 9.65 10.73 9.39 8.52 Total Direct Taxes 124.73 118.24 140.04 138.50 130.58 112.19 79.53 Trend of Direct Taxes 2/ 139.85 132.43 125.41 118.76 112.L 6 106.51 100.87 Indirect Taxes on Consumption 328.4 335.7 389.0 465.5 447.4 550.5 548.4 Trend of Indirect Taxes 2/ 322.7 355.0 390.48 429.5 L72.4 519.7 571.6 - In percent of agricultural output (gross value added) at constant (1964/65) prices - Land Revenue 1.22 1.13 1.39 1.35 1.15 1.15 0.80 Agricultural Income Tax 0.12 0.11 0.11 0.10 0.10 0.11 0.10 Total Direct Taxes 1.34 1.24 1.50 1.45 1.25 1.26 0.90 Indirect Taxes on Consumption 3.5 3.5 4.2 h.9 4h.3 6.2 6.2 1/ Deflator: Index of Investment 2 cost at 1964/65 prices, Plan- RI ning Commission. 86 89 93 96 100 109 124 2/ Least squares estimate of 'b" in the equation Log Y = a + b.time. Source: Ved P. Gandhi, Taxes, SubsidieS and AgriCUltUral PrOdUCtion (India). Ved P. Gandhi, Tax Burden on Indian Agriculture, Harvard Law School Tables 12 and 13. Rprt on Currency and Finance 196667, Statements 52 and 56. Table 4: mpen and Concealed Subsidies to Indian Agriculture Rs. in million 1963/64 1964/65 1965/66 1966/67 1967/68 - in current prices - Open Subsidies f or: Fertilizer and Manure 22.7 30.0 L2.9 18.5 17.1 Seeds 7.1 8.1 9.1 11.4 12.1 Pesticides 5.6 7.2 9.7 12.6 13.3 Other imputs 33.7 38.6 42.6 57.4 48.3 Sub-total 69.1 83.9 lo4.3 99.9. 90.8 Concealed Subsidies for: Irr"igation 324 366 388 415 Electricity 6 6 6 6 10 Fertilizer 15 407 Tacari Loans 9.8 9.4 9.4 9.8 Co-operative Credit 29.39 30.21 36.18 38.85 Sub-total 369.19 411.61 454.58 876.65 Total Subsidies 438.29 495.51 558.88 976.55 Total Subsidies (constant 1964/65 pricesl/) 456*55 495.51 512.73 787.54 1/ Deflated by Index of investment Cost, Source: Planning Commission. Source: Ved P. Gandhi, Taxes, Subsidies and Agricultural Production (India), Indian Institute of Managernent, Ahmnedabad, September 96-b