Distributional consequences of alternative food policies in India

DISCUSSION PAPER Report No.: ARU 20 Distributional Consequences of Alternative Food Policies in India Hans P. Binswanger and Jaime B. Quizon Research Unit Agriculture and Rural Development Department Operational Policy Staff World Bank August 31, 1934 The views presented here are those of the author(s), and they should not be interpreted as reflecting those of the World Bank, The authors are staff member and the consultant of the World Bank. However, the World Bank does not accept responsibility for the views expressed herein which are those of the authors and should not be attributed to the World Bank or to its affiliated organizations. The findings, interpretations, and conclusions are the results of research supported in part by the Bank; they do not necessarily represent official policy of the Bank. The designations employed and the presentation of material in this document are solely for the convenience of the reader and do not imply the expression of any opinion whatsoever on the part of the World Bank or its affiliates concerning the legal status of any country, territory, area or of its authorities, or concerning the delimitation of its boundaries, or national affiliation. DISTRIBUTIONAL CONSEQUENCES OF aLTERNATIVE FOOD POLICIES IN INDIA Hans P. Binswanger and Jaime B. Quizon A number of large and poor economies (India, China, Bangladesh) are not open economies but economies where the state controls trade in food and also in many non-food items. Food policy interventions such as food aid, supply oriented interventions or ration shops alter the balance of supply and demand for food and can therefore be expected to affect food prices unless they are accompanied by compensating variations in imports or exports. In this paper we investigate various food policy options in India, considering both their direct and indirect effects on consumption via prices and real incomes. In considering these food policy options one is usually interested in food consumption of the poorer groups. While showing the impact of the policies on cereal consumption of the poor, we will concentrate the main attention on the real income consequences of these policies. First, food consumption and real income are highly correlated. Moreover it is now widely accepted [Reutlinger and Selowsky (1976), Sen (1981)] that lack of food security is more a matter of lacK of income of the poor to purchase food than of a general shortfal.1 in food supply. In considering food security issues it is the income effect of food policies, not the supply effects, which are of primary importance. This does not, of course, distract from the importance of supply issues in other contexts. -2- In much of this paper we will discuss various forms of food subsidies to be contrasted at the end with supply oriented policies. Food subsidy policies are policy packages which - implicitly or explicitly - imply decisions on (or consequences for) at least three policy dimensions. (1) Targeting Who is to receive the subsidy? The subsidy may be untargeted, i.e., simply be reflected in a lower open market price, or it can be targeted to specific groups. In this paper we consider two options. Urban ration shops target the subsidy to all urban groups. The alternative option is a food stamp scheme which targets the subsidy to all poor, regardless of place of residence. (2) Procurement Where do additional supplies come from? The extra demand for food can be accommodated by extra imports, in which case there will not be a price effect. On the other hand if imports are not increased, the extra demand must come from domestic producers who will be unwilling to supply more food unless prices rise. (3) Finance How is the food subsidy financed? We consider three options: foreign aid, an excise tax on non-agricultural consumer goods and a forced levy of foodgrains from the two largest farm size groups. Other ways of financing not considered here include income taxes, inflationary deficit financing and reductions in other government expenditures. In discussing food subsidy issues it is sometimes argued that finance issues should be ignored as it is often not possible to match the food subsidy with a specific source of finance. For example supression of a food subsidy could lead the government to engage in some other form of expenditures whose direct and general equilibrium impact cannot be evaluated. While this is undoubtedly the case, it is possible to show the -3- implications of various financing schemes, each with sharply different initial incidence. As we will show below, alternative financing schemes have vastly different implications, reinforcing or offsetting the distributional objectives of the food subsidy. In the general equilibrium model of India's agricultural sector which'we will present, we will consider all three dimensions explicitly. However, there are at least three additional effects which we are not able to capture. These are the impacts of food policies on the foreign exchange market, on non-agricultural employment and wages and on private investment. We will consider these issues informally in Section 4, where we also discuss some studies which are able to incorporate them formally. 1. A Brief Overview of the Model The model which we use has been described in great detail in Quizon and Binswanger (1984). It is a general equilibrium model of the agricultural economy of India in which agricultural production, prices and incomes are determined endogenously for given levels of non-agricultural production and nominal non-agricultural incomes. Real incomes, however, are endogenously determined as nominal incomes, deflated by a price index which depends on food prices. Total demand for agricultural commodities reflects the real incomes of all groups, i.e., the food price effects of policies affect demand via their impact on both rural and urban consumers. Non-agricultural commodities are the numeraire and the model does not endogenously take account of exchange rates and trade issues or of monetary phenomena. -4- The elements of the model are as follows: 1. Producer core: A system of output supply and factor demand equations describes producer behavior in each agroclimatic region. It determines aggregate supply of: - rice - wheat - coarse cereals - other agricultural commodities It determines aggregate demands for variable factors: - labor - draft power - fertilizer A variety of shifter variables may shift each of the supply and demand curves. They are: - land, a fixed factor - rainfall - irrigation - high yielding varieites (HYV) - roads - farm capital - regulated markets - technological change The parameters of this system have been econometrically estimated for each agroclimatic region and aggregated up to obtain the All-India parameters. A flexible functional form has been used and all across price terms, including those between inputs and outputs, have been estimated. No separability restrictions have been used. The agroclimatic regions are: (a) northern wheat region (b) eastern rice region (c) coastal rice regions of south India (d) semi-arid tropics 2. Income Groups are defined as quartiles of the All-India rural and urban expenditure distributions respectively, with R1 and Ul being the lowest rural and urban quartiles, respectively, and R4 and U4 the highest. - 5 - 3. Input Supplies to agriculture are determined as follows: land: exogeneously given, supplied only by each of the rural groups labor: responsive to the real rural wage and supplied by each of the rural groups and by urban groups via migrationl! draft power: responsive to real draft animal rental rates and supplied by each of the rural groups fertilizer: aggregate supply curve responsive to the price of fertilizer relative to non-agricultural goods. 4. Consumer core: output demands: Commodities are demanded by each of the eight groups according to the prices of the commodities and according to their real income. Each group's demand is modeled separately according to income group specific demand elasticities, i.e., poorer groups have higher income elasticities than richer groups. The demand systems have been estimated econometrically. A flexible functional form has been used so that all Slutzky substitution terms have been directly estimated. Aggregate demand is the sum of each group's demand. 5. Nominal income of each group is computed as their respective supplies of factors of production times the factor prices, plus an exogenously given component of non-agricultural income. 6. Real income is defined as nominal income, deflated by an endogenous consumer price index. The price index reflects all endogenous food price changes and is specific to each income group. 7.. Price and quantity determination: Prices and quantities are determined as that which equate aggregate demands with aggregate supplies for each of the four agricultural outputs and the three variable inputs. The quantity of land is exogenous, but the "land rent" is determined endogenously as the 1/ As non-agricultural output is given, changes in migration will create excess demand or supply of labor in the urban sector, or unemployment. -6- residual farm profits after variable factors have been paid off. Non-agricultural prices are given exogenously, quantity consumed adjusts.2/ 8. The model solves for these prices and quantity changes simultaneously determines for each income group the change in: - nominal income - price deflator - real income - labor and draft power supply - consumption levels. The solution is simple because the entire model is-written in logarithmic-linear form. 9. Base-year initalizing quantities, prices and shares: The base year is 1973/74 and the initializing values have been computed largely from the NCAER/ARIS survey [for details see Pal and Quizon (1984)]. 2. Food Subsidies in a State Trading Economy In Tables 1 and 2 we report the likely effects of particular government intervention schemes on various economic indicators of general concern to policymakers. The numbers herein show the percentage change of a particular indicator (say, real national per capita income) given a selected intervention scheme from an underlying level determined by a reference growth path with-no intervention. This is better explained by Figure 1. 2/ As non-agricultural income is given exogenously, non-agricultural production is also exogenously given, i.e., consumption must adjust via trade. Figure 1 t t+1 Time In Figure 1)let A describe the normal path of an indicator which would occur in the absence of specific food policy intervention. Let B be the path which would obtain with the intervention. The value c is then the difference between the levels of the variable defined by paths A and B at some period (t+1), i.e., c = b-a. The ratio c/a is the number reported in Tables 1 and 2 for any indicator, i.e., it is the percentage difference in the indicator caused by the policy intervention. At the mcment we do not model the underlying reference growth path of the economy. Nor do we consider the administrative costs or the political risks associated with any particular government scheme. These costs and risks may differ widely across schemes. -8- Recall now from the introduction that a food subsidy involves explicit or implicit decisions on targeting, imports and finance. Scenarios 1 and 2 in Table 1 are untargeted food subsidies. US$1.1 billion are used to finance the import of almost seven million tons of wheat which are released in the open market. In scenario I these imports are aid-financed and add about 0.5% to real national income. In scenario 2 the imports are financed by a domestic excise tax on household consumption of non-agricultural commodities of about 8.5% which leads to a decline in disposable income of 2.5%. Both scenarios sharply reduce the wheat price (-60%). As other agricultural commodities are substitutes of wheat, their prices also decline. The GNP deflator declines by about 19% . Wheat production declines by around 18%. The decline in rice production is minimal while coarse cereal production and other crop production actually increase modestly. Thus, aggregate agricultural supply declines by only about 0.6%. This small decline in output leads to very small increase in agricultural employment (less than 0.1%) but somewhat larger declines in real rural wages of between 1.0 and 1.6%. The agricultural price declines sharply affect residual farm profits (-34% and -39%). The big gainers in both scenarios are the urban poor whose real incomes rise by about 18%. The other urban groups gain as well but because they spend proportionally less on food, their relative gains are smaller. The rural poor, despite their loss in wages, gain by about 4-5%. This arises because food price declines dominate the negative wage effect. But members of the second rural quartile are already net losers, as the sharp decline in farm profits outweigh their consumption gains from the food price drop. - 8a - Table 1: UNTARGEIED FOOD STBSIDIES AND URBAN RATION SHOPS Assumvtion (1) (2) (3) (4) (5) (6) Targeting None None Urban Ration Urban Ration Urban Ration Urban Ration Shops Shops Shops Shops Trade Foreign Foreign Foreign Foreign Domestic Domestic Supply Supply Supply Supply Supply Supply Finance Aid Excise Tax Aid Excise Tax Excise Tax Forced Procurement Variables REAL NAT.PER_CAP INC. -0.175 -2.513 1.728 0.282 0.4719 0.3934 TOTAL OUTPUT -0.624 -0.622 -0.171 -0.169 0.1930 0.0835 J OF RICE PRODUCED -0.503 -3.741 -0.388 -0.532 0.3065 0.0282 mHEAT PRODUCED -L7.318 -19.120 -4.123 -5.238 0.7906 0.4216 CC PRODUCED 8.018 9.978 1.517 2.729 -0.6478 -0.1485 ut PRUDUCED 2.063 2.090 L.027 L.044 0.2508 0.1377 5NP DEFLATOR -20.154 -18.236 -2.726 -1.539 0.6191 2.2664 PRICES OF RICE -29.209 -29.552 -4.9L8 -5.130 7.3881 2.9924 mHEAT -59.036 -62.106 -LL.863 -13.761 8.0434 3.5874 CGARSE CEREALS -21.066 -18.224 -2.909 -1.152 7.1634 3.0827 OTHER CROPS -17.831 -18.113 -0.684 -0.858 6.b238 2.9215 REAL dAGE RATE -1.090 -1.565 -0.505 -0.799 -0.3880 0.1155 LABOR EMPLOYMENT 3.099 0.019 -0.090 -0.139 -0.1984 -0.0077 REAL RESIDUAL PROFITS -34.413 -38.814 -3.043 -5.764 8.6805 4.684L REAL PER CAP.INC.RURAL L 5.189 4.311 0.864 0.321 -1.8233 -0.5182 RURAL 2 -0.194 -L.564 0.328 -0.519 -0.3716 0.2363 RURAL 3 -3.922 -5.779 -0.054 -1.203 0.5050 -0.9027 RURAL 4 -11.466 -14.542 -0.873 -2.774 2.1849 -1.5439 URBAN 1 18.862 17.819 L2.652 12.007 4.5240 8.0958 URBAN 2 18.330 16.744 10.372 9.391 1.8657 6.0317 JRBAN 3 16.054 13.807 7.392 6.002 -0.5453 3.6917 URBAN 4 9.025 5.050 3.763 1.305 -2.4027 1.9718 PER CAP.CER CONSUME RI 6.153 5.830 1.686 1.487 -1.2159 -0.5056 RURAL 4 0.700 1.246 1.465 1.302 1.7766 -0.3170 URBAN 1 20.532 20.079 L1.294 LL.014 30887 5.7458 URBAN 4 0.973 0.940 L.841 1.821 1.2374 1.1470 - 8b - :able 2: FOOD STAMPS TO ALL POOR AND SUPPLY ORIEE D PROGRAMS Assu=otion (7) (8) (9) (10) (11) (12) (13) (14) Irrigation Fertilizer Technical Change Investment Subsidy largecing All Poor All Poor All Poor All Poor None None None 'one Trads Foreign Foreign Domstic Domestic No Exports No Exports No Exports With Supply Supply Supply Supply Exports Finance Aid Excise Tax Excise Tax Forced Pro- Aid Aid Aid Aid curement Variables REAL NAT.PER-CAP [NC. 4.027 -O.J33 3.497 0.2727 1.712 1.30L0 7.197 L.1499 TOTAL OUTPUT -.370 -0.361 3.652 O.j436 2.287 0.6565 9.6L2 11.541 ;F RICE PvOOUCED -0.343 -0.739 1.608 0.32+4 0.636 0.341.2 10.347 13.84 ,HEAT PRODUCED -11.615 -14.od-, 2.195 L.154+ 5.140 1.2905 7.493 21.11.3 CC PRODUCED 4.13L 7.502 -1.951 -0.54.41 1.878 -2.132Z 7.276 -0.402 ZC PRODUCED 2.308 2.347 J.127 -.L906 3.482 2.464 11.003 10.85a -AP 3EFLATOR -9.156 -5.737 17.105 -t.AoS2 -5.756 -1.140 -18.125 22.4 PRICES OF RICE -14.512 -L4.977 20.070 7.7357 -6.926 -1.7581 -25.948 32.417 "HEAT -35.033 -40.138 20.924 8.3725 -12.771 -1.7814 -31.236 44.46 GUARSE CE,EALS -10.095 -5.070 18.211 6.7375 -9.387 1.2509 -33.097 16.879 j 3THER CRCPS -.823 -5.233 15.7L,* 5.3000 -6.j7b -L.938L -19.550 2o.587 REAL .AGE RATE -1.215 -2.337 -0.Sd7 0.525 0.705 -1.399s -0.097 2.22+ LABOR EMPLOYMENT -0.173 -0.313 -3.479 0.0578 0.438 -0.7691 0.592 0.463 ,EAL RESIDUAL PRCFITS -12.351 -09.846 20.594 9.3473 -4.792 5.5813 -4.309 74.742 ,EAL PER CAP.INC.RURAL 1 17.618 L6.069 10.066 13.7351 2.917 -0.3536 9.945 0.067 RURAL 2 10.662 8.24 8.o97 1).4048 1.714 0.7526 7.258 10.233 RURAL 3 -0.853 -4.061 0.719 -3.2430 0.397 1.5446 5.674 17.01 RURAL 4 -3.775 -9.068 4.817 -5.6751 -0.674 2.5395 1.836 30.534 JRBAN 1 20.055 18.163 -2.787 7.2576 6.040 0.5977 17.779 -16.0,5 URBAN 2 16.939 14.104 -6.965 4.750 5.728 0.7436 17.505 -19.501 URBAN 3 6.648 2.674 -15.658 -3.7443 5.154 0.6025 15.190 -17.201 URBAN 4 2.820 -4.1.26 -14.507 -2.2087 3.503 0.3991 9.832 -10.533 PER CAP.CcR CONSUME RI Li.713 13.121# 5.557 7.5545 2.570 -0.7351 11.143 -0.531 RURAL 4 2.932 3.881 3,808 -2.0793 -0.071 0.5953 L.176 9.246 UK8AN 1 19.986 19.116 -3.074 4.3987 5.591 0.0649 16.723 -14.12Z UASAN 4 1.755 1.b92 3.059 -0.1955 -0.382 -0.3330 0.386 1.272 -9- The next four scenarios provide food subsidies to urban vesidents only via ration shops. These shops provide each member of the urban population with the same amount of subsidized food, irrespective of their income level. The amounts are set such that the poorest urban group obtains an initial nominal income boost of 10%. We further assume that the food rations are either inframarginal, or if not, the excess over one's own demand can be easily resold in the open market. This allows us to model each scenario simply as a nominal income transfer. In scenarios 3 and 4 we repeat the import and financial arrangements of scenarios 1 and 2. The supplies for the fair price shops are imported. The imports are aid-financed in scenario 3 and excise tax-financed in scenario 4. Both-wheat and rice are imported but as wheat is much cheaper internationally than rice, imports concentrate on wheat. Two million tons of wheat and 0.8 million tons of rice are imported for a total of US$560 million. We then shift the source of suplies to the domestic producer in the next two scenarios. In scenario 5 the same excise tax of 5.25% levied in scenario 4 is used to procure the supplies domestically, while in scenario 6 the supplies come from forced procurement from the two richest rural groups, with rural 4 having to contribute twice the quantity of rural group 3. These procurement taxes amount to 1.63% and 3.26% of the nominal incomes of these two group, When the supplies for the ration shops are imported, the urban poor experience a double gain, i.e., the initial boost in income of 10% an the effect of lower food prices on the GNP deflator of 2.7% and 1.5%, respectively in scenarios 3 and 4. Since the urban poor spend little on non-agricultural commodities, the excise tax has only a minimal impact on them. - 10 - On the other hand when the extra demand of the urban groups is not accommodated by imports, food prices rise, leading to increases in the GDP deflator of 6.6% and 2.3% respectively in scenarios 5 and 6, and a corresponding erosion of the initial income gains of the urban poor. The urban rich, who spend relatively little on food and much on non-agricultural commodities, are affected as much or more by the excise taxes than the food price effects. They gain about 4% of real income with aid financed food imports, but lose by 2.4% with excise tax financed domestic supplies. The real income of the rural poor is relatively insensitive to the urban ration shops. They gain marginally under both import scenarios but lose marginally when food is domestically procured.- The increases in domestic supply do not have a sufficient labor demand effect to offset the impact of higher food prices on their real income. The rural rich lose when supplies for ration shops come from imports. But they also lose when the supplies come from forced procurement. Increased demand from urban consumers does not raise the open market prices sufficiently to compensate the rural rich for the targeted tax which is implicit in the forced procurement. This finding is contrary to the assertion of several authors that forced procurement is not necessarily a disincentive to rich producers because the weighted average price realized by farmers may not be lower than the equilibrium price in the absence of procurement [Dantwala (1967), Mellor (1968), Hayami, Subbarao and Ostuka (1982)]. The rural rich can only experience a gain if the urban fair price shops are financed by an excise tax or another tax with a more general incidence than forced procurement. - 11 - Of the four urban ration shop schemes, domestic procurement with excise tax is by far the least effective in raising the real income of the urban poor. Under this scheme real incomes of the poorest urban quartile increase by less than 5% compared to the 8-13% increases in the three other plans. On the other hand excise tax-financed imports neither largely subsidize the urban rich nor heavily penalize the rural rich. In Table 2 we explore an alternative program which provides food stamps or food rations to the poorer half of the population only, namely the rural and urban groups 1 and 2. The food stamps are scaled such that the poorest rural group obtains an initial boost in its nominal income of 15%. This translates into gains in nominal income of 10% for rural 2, 11% for urban 1 and 9% for urban 2. We explore the same four procurement and financing scenarios as with the urban ration shops. The poverty oriented program is much larger in magnitude than the urban fair price shops as half of the population now receives food aid. (The total urban population eligible in the previous scenarios is only 20% of India's population). To provide for the expanded program by imports, 6 million tons of wheat and 2.1 million tons of rice would have to be imported. The excise tax rate would have to be 14.75% of household consumption of non-agricultural goods; the procurement taxes would be 4.6% of the income of rural group 3, and 9.2% of the income of rural group 4. As the programs are large and transfer income only to low income groups, food demand effects are very large. When these increases in food demand are accommodated by imports (scenarios 7 and 8), the beneficiaries of rations gain both from price declines and the rations, with the gains of the urban poor exceeding those of the rural poor. As domestic food production declines, the rural poor lose some employment and some wages. - 12 - With foreign aid the rural rich lose about 3.8% of their real income while losses of group R3 are insignificant. With imports financed via excise taxation, an excise tax burden is added to the farm profit losses and the real burden of the scheme falls on rural group 4 (-9%), rural group 3 (-4%) and urban group 4 (-4.1%). With domestic procurement the added food demand leads to food price increases. These are especially large under the excise taxation scheme where the GNP deflator rises by 17% compared to a 5% increase under forced procurement. With excise taxes the price rises more than offset the value of the food ration for urban groups I and 2 who lose between 3% and 7%, i.e., two of the target groups end up at net losers rather than gainers. Thus the program does not meet one of its distributional objectives. The major burden of the scheme is borne by the urban groups 3 and 4 who lose about 15% of real income. On the other hand all rural groups gain. But the gains are unevenly distributed. Rural group 3, which does not receive a ration, just barely stays even. The effect of food prices on profits just dominates the excise tax effect and the price effects on the price index of rural groups 3. Food stamps with forced procurement achieve the distributional objectives much better. All four target groups show a real income gain. Large farmers bear the heaviest burden followed by rural group 3, urban group 3 and urban group 4. Employment and wage effects are minimal. 3. Supply Oriented Interventions In scenarios 11 to 14 we consider supply-oriented intenventions. These are an irrigation investment, a fertilizer subsidy and a technical change. As these interventions affect food prices, they amount to - 13 - untargeted "food subsidies". Furthermore, we do not as of yet have good cost estimates for these interventions and therefore assume that they are aid-financed, making them comparable to scenario 1. In scenarios 11 to 13 we assume that additional supplies will not result in additional exports (or reduced imports), i.e., we force domestic absorp:ion of all extra supplies. In scenario 14 on the other hand, we make the opposite and extreme assumption that all the additional supplies made feasible by the technical change are used to reduce imports or add to exports.3/ As we have no cost comparisons of the three supply-oriented policies, only the direction of change should be considered not the magnitude of the effects. The irrigation investment is assumed to increase irrigated area by 10%. The fertilizer subsidy is assumed to initially increase fertilizer prices received by domestic producers by 40%, thereby initially increasing available fertilizer by 80%. In the technical change scenarioI yields per hectare of all crops are assumed to be increased by 10%. All these supply-oriented policies tend to reduce agricultural prices when domestic absorption is forced. Wheat production grows most when irrigation is increased while other crops grow most when fertilizers are subsidized. The price declines clearly benefit the urban consumers and especially the urban poor. The rural poor gain both from irrigation investments and from technical change. Thus, both these supply-oriented programs can be used as substitutes for subsidies. However, the fertilizer 3/ Technical change has three effects. The initial shock is the increase in yield made possible by technical change, holding input levels constant (we assume that this initial shock is available for exports or reduced imports. At constant prices farmers will then find that relative profitabilities have changed and adjust their crop mixes. This is the secondary effect. The initial shock and the second effect combined then have a third set of general equilibrium effects, e.g., they may depress food prices and lead to further adjustments in the supply of farmers. All three effects are traced in the model. - 14- subsidy does not benefit the rural poor at all. Fertilizers are substitutes for labor either directly and/or indirectly by changing the crop mix in favor of less labor intensive but fertilizer responsive commodities. Therefore, they lead to real employment or wage declines whose negative effect on the nominal incomes of rural poor exceed 'the positive food price effects. Thus fertilizer subsidies do not appear to be a good substitute for food subsidies. Contrary to most people's intuition, irrigation expansion can lead to real income losses for the rural rich (as long as all extra supplies are domestically consumed) because the added supplies depress prices without otherwise reducing costs of production. Technical change is more beneficial to large farmers as food price effects are partly offset by the cost reduction associated with tehcnical change. Some, but not all, agricultural supply interventions can have beneficial income and food consumption effects for the poor, provided that the added supplies are used to drive down domestic prices. Scenario 14 sharply highlights this cruci-al proviso. If the initial shock of the technical change (its yield effect at unchanged input levels and output prices) is used to reduce imports or increase exports, the income distribution effects are completely reversed. Because the extra income made possible from the technical change adds to the demand for food, prices now rise rather than fall. Thus the urban poor are big losers, followed by all other urban groups in the now familiar manner. The rural poor neither gain nor lose as their losses from food price rises are compensated via employment, wage and small farm profits effects. The rural rich are the largest gainers. Production gains from irrigation and fertilizer subsidy could also be exported. This trade policy is the largest determinant of - 15 - the distributional effects of the supply-oriented policy. Moreover, taxation issues, not considered here, can further alter the distributional outcomes. The two technical change scenarios represent extreme cases of trade policy, i.e., complete domestic absorption or complete exports of the initial gains. The reader can consider intermediate cases easily. For example the consequences of exports of one-half of'the intial shock can be computed as the simple average of the two scenarios. India's historical experience with the Green Revolution was also such a mixed case. The extra supplies were first used to reduce dependence on imports, but did not lead to extra exports once self-sufficiency was achieved. 4. Other Potential Effects of Food Policies There are several effects of these food policy packages which we are not able to consider in our model. The most important are the repercussions of the food policies on (a) the market for foreign exchange; (b) employment in non-agricultural sectors of the economy arising from changes in final demand; (c) real wages of urban workers arising from the effects of changes in rural labor demands on urban labor markets; and (d) investment. In the absence of a formal model which explicitly takes these effects into account, it is not possible to evaluate their importance. However, a few general points and cautionary remarks are in order. Foreign exchange markets: When food policies rely on domestic procurement or aid financed imports, the balance of supply and demand for foreign exchange is not directly affected. Thus, many of the scenarios considered have no direct foreign exchange effect. With aid-financed imports this is because the foreign aid provides the foreign exchange which - 16 - buys the extra food.4/ Only indirect effects via changes in demand for tradeable non-agricultural commodities arise, and, as we shall see below, these effects are hard to predict. When extra food is procured internationally but is financed from domestic sources as in scenarios 2, 4 and 8, the extra demand for foreign exchange can be accommodated in several ways, greatly complicating the analysis: by a decline in the value of the domestic currency; by a reduction in foreign reserves; by a reduction in non-agricultural imports or by a combination of these three. Scobie (1981), for example, shows that Egypt has attempted to stabilize the exchange rate and accommodated the extra food imports primarily by reductions in non-agricultural imports, and by an increase in the instability of such imports. As our model does not trace year to year fluctuations, such instability effects cannot be considered. Also, if one is concerned with long run impacts of a commitment to food subsidies and higher import levels, it is best to consider a case where the exchange rate adjusts. A downward adjustment in the value of the rupee arising from increased food imports will tend to increase foreign demand for non-agricultural tradeable goods. Whether that will lead to an increase in employment in the non-agricultural sector, however, is not clear, as the excise tax required to finance the extra imports will dampen the demand for all non-agricultural commodities (tradeable and non-tradeable). In the absence of a formal model we cannot therefore predict the long run impact on the economy via the exchange rate. 4/ Note that if aid is transitory but food subsidies cannot be withdrawn when aid stops, the policies will have delayed foreign exchange effects. - 17 - Non-agricultural labor demand: When food is domestically procured) there is no direct foreign exchange effect of food policies. But there will still be changes in non-agricultural employment which are not traced in the model. These effects are also hard to predict. Consider scenario 9, food stamps to all the poor financed by a domestic excise tax. As food prices increase, there is a price effect in favor of non-agricultural commodities. But excise taxes partly offset this effect. Moreover, the excise tax-financed food policy scenario reduces the incomes of the rich urban groups substantially, while it raises that of the richer rural groups only very modestly. Thus the income effect is likely to be negative. The net effect on non-agricultural demand (and hence on non-agricultural employment) can only be measured with a formal model which treats the non-agricultural sector endogenously. Impact of agricultural employment on urban wages: Our model allows us to predict agricultural labor market consequences. Changes in excess demand for agricultural labor will spill over into the non-agricultural sector via migration. In scenarios 2 and 8, for example, agricultural wages and employment both decline. Such declines should reduce the wages in the non-agricultural informal sector sooner or later, reducing the benefits which the urban poor derive from these policies. However, our model shows that agricultural employment and real wage effects of the various food policies are generally quite small, so it is unlikely that this labor market spillover, if incorporated, should substantially affect the conclusions of this paper. A study by Amranand and Grais (1983) on Thailand, however, shows that such a conclusion is country-specific (or perhaps model-specific). The Thai model is a fairly disaggregated general equilibrium model. The market for casual labor (agricultural and non-agricultural) is treated in a - 18 - neoclassical fashion as adjusting to supply and demand. Formal sector wages are sticky downwards within each year, but adjust slowly downward over several years if supply exceeds demand for a prolonged period. Total investment and its allocation to sectors is endogenous. The exchange rate is managed. Changes in agricultural trade are accommodated by changes in exchange rate reserves and changes in foreign borrowings. The study shows the effect of a 5% increase in rice export tax during the first year of the tax and after six years.5/ The tax revenue is :ssumed to reduce the government deficit, i.e., more credit is available for other investors. The 5% tax reduces the domestic rice price by 4.2% which translates to a real income gain of 0.8% for casual workers (both rural and urban) and rubber farmers. Gains of blue collar and white collar workers are small and crop farmers lose by nearly 2%. The short run employment effects are minimal because outputs do not adjust within a single year. In the longer run)depressed agricultural prices lead to a decline in aggregate agricultural output of 0.5%. As rice is labor intensive, aggregate labor demand declines by 0.25%. The study does not show the long run wage decline but it must be about 1% because real casual labor income drops by 0.2% in the long run compared to the short-term rise of 0.8%. Crop farmers must also be losers in the long run, as the tax will continue to depress prices. Since many rice farmers are poor, the general equilibrium results show that the cheap food policy via the export tax has no long run distributional benefits in the Thai case. The difference with the India results arise probably because rice, the crop whose price is depressed, is very labor intensive compared to wheat and because the foreign trade in rice is very large and price elastic. 5/ They simulated a reduction in the export tax. We simply changed the signs on their results. The results refer to their simulations with a foreign demand elasticity on rice of -3.0. - 19 - Investment: While agricultural investment is not endogenously treated in our model, the fact that residual farm profits are explicitly shown allows us to predict what would happen to private farm investment. Food policies which rely on domestic procurement will tend to raise farm investments while those which rely on imports will tend to depress it. Thus, food policies which in the short run look relatively unfavorable because rising prices'erode some of their benefita may be more attractive in the long run when private farm investments lead to shifts in supply and reductions in prices. And the reverse holds for policies which initially lower prices. We are not able at this time to show how important these long run consequences are. No econometric studies exist which would enable the prediction of the extra investments which would be forthcoming when profits rise. The impact of the investments themselves onthe supplies are more easily evaluated. That the investment issues are potentially important is, however, demonstrated by a study of Argentina by Cavallo and Mundlak (1982). They built a highly aggregative dynamic two-sector model and estimated both its static and dynamic parameters econometrically. The labor market is treated in a neoclassical fashion but with rural to urban migration responding sluggishly to changes in demand and supply. Cavallo and Mundlak's model simulates the period 1950 to 1971. They first show that trade liberalization would only have benefitted Argentina if it had been accompanied by exchange rate management aimed at keeping the value of the peso below a free market equilibrium. However, they observe that such a scenario results in reductions in the urban wage relative to food prices compared to their baserun. They therefore explore a scenario in which trade liberalization cum exchange rate management is accompanied by a subsidy on agricultural output. This is implemented by - 20 - reducing the domestic tax on agricultural output while increasing the domestic tax on non-agricultural output to hold government revenue constant. We compare short and long run consequences of the scenario with food subsidy to the scenario with only trade liberalization and exchange rate management.6/ In the short run (2-3 years) the subsidy increases the food wages in both agricultural (urban) and non-agricultural (rural) sectors. Because of the increased tax on non-agricultural goods it must, however, decrease the wage in terms on non-agricultural commodities. Real wages therefore would increase only for the very poor who spend a large proportion of income on food. This group has always been much smaller in Argentina than in India. In addition there are very small short run reductions in per capita income, per capita consumption and investment. In the long run) the cumulative effect of reduced investment in both the agricultural and non-agricultural sectors dominates all other effects of the subsidy. After 20 years most relevant macro-aggregates are lower than without the food subsidy: per capita income (- 13%), per capita consumption (- 12%), investment (- 4%), agricultural production (- 4%). In addition urban workers are unambiguously worse off, as they have both lower food and non-food wages. Rural worker's food wage increases by 2% while their non-food wage declines by 1%, i.e., their real wage is largely unaffected. Thus the short run benefits of low food prices are not impressive and the long run effects unfavorable. These results are of course sp-cific to Argentina. In India the short run distributional benefits of low food prices are substantially larger as a much larger portion of the population is very poor, i.e., spends the largest share of its income on food. 6/ We do not compare the subidy scenario to Cavallo and Mundlak's base run, a comparison which would be inappropriate as other things are not held constant. - 21 - 5. Conclusions The discussion of the previous section shows that our modelling effort has omitted a number of considerations which might modify some of the conclusions reached in Tables 1 and 2. Most importantly we are not yet able to incorporate the long run effects of food subsidy policies on private agricultural investments. Thus the conclusions should be regarded as most relevant to the intermediate run of one to three or four years. The food subsidy scenarios considered clearly show the importance of evaluating the entire policy package associated with the food subsidy, including the trade and taxation aspects. Ignoring the price effects of the trade decisions or the income effects of taxation can easily result in highly misleading conclusions on the distributional impacts of the subsidies. We show several cases where the trade and taxation measures lead to a complete erosion or even reversal of the intended real income effect of the food subsidies. As real income and food consumption effects are closely related, nutritional objectives of food policies may also not be acheived. More specifically)we find the following: 1. Untargeted food subsidies achiev'ed via additional imports benefit both rural and urban poor regardless of whether they are aid or excise tax financed. Compared to aid-finance, excise tax-finance hurts both urban and rural rich. The excise tax ends up being progressive as poor groups spend relatively little on the taxed non-agricultural commodities. 2. Targeted food subsidies can achieve their stated objectives much more easily if the additional supplies required by the demand increases are imported than if they are not. In extreme cases)some of the target groups may become net losers if the food rations are not imported. More - 22 - generally, the rations required to achieve a specific distributional objective are much smaller when the rations are imported and added to supplies than when they are not. 3. There is likely to be intense conflict about the import policies. The rural rich, the net sellers of food, benefit from preventing imports or increasing exports regardless of the financing or targeting scheme. The rural poor and all urban groups, i.e., the net buyers of food, tend to benefit from the opposite policies. The conflict is then not between the urban and the rural groups but instead between all net buyers and all net sellers. 4. Taxation issues do not only involve the potentially taxed groups. A shift of the tax burden from the rural rich (as under forced procurement) to all groups (as under the excise tax) has general equilibrium effects which affect incomes of the target groups via food prices. Taxing the rural rich, the largest food consumers, reduces aggregate food demand by more than taxing all groups via an excise tax. Therefore, the poor target groups of the food subsidy benefit more from the forced procurement scheme than from the excise tax-financed scheme. 5. All food subsidy schemes have agricultural employment and wage effects which are small compared to the gains and losses incurred by the rural poor as a result of the price effects of a food subsidy or ration shop policy. (We have not considered urban and/or formal sector wages). They are also small compared to the effects of rations directly allocated to the rural poor. Domestic procurement for food subsidies and rations tends to slightly increase employment and wages, as domestic supplies increase, while foreign procurement (whether aid or tax-financed) discourages domestic production and leads to corresponding employment and wage losses. - 23 - 6. Some supply oriented policies can achieve similar distributional objectives as food subsidies. But because they must achieve these objectives via reduced market prices, they can do so only if the economy is closed, i.e., if the extra supplies do not result in increased exports or reduced imports. Moreover, not all supply oriented policies benefit the poor. A fertilizer subsidy does increase output supply but tends to erode employment and wages while supporting farm profits. -24- REFERENCES Amranand, Piyasvasti and Grais, Wafik. Macroeconomic and Distributional Implications of Sectoral Policy Interventions: The Case of Energy and Rice in Thailand. World Bank Staff Working Paper 627. Washington, DC: World Bank, 1984. Cavallo, Domingo and Mundlak, Yair. Agriculture and Economic Growth in an Open Economy: The Case of Argentina. Research Report 36. Washington, DC: International Food Policy Research Institute, 1982. Dantwala, M.L. "Incentives and Disincentives in Indian Agriculture". Indian Journal of Agricultural Economics, 22 (1967): 1-25. Hayami, Yujiro; Subbarao, K.; Otsuka, Keijiro. "Efficiency and Equity in the Producer Levy of India". American Journal of Agricultural Economics. (November 1982): 655-663. Knudsen, O.K. and Scandizzo, P.L. Nutrition and Food Needs in Developing Countries. World Bank Staff Working Paper 328. World Bank: Washington, DC, 1979 Mellor, John W. "Functions of Agricultural Prices in Economic Development" Indian Journal of Agricultural Economics 23 (1968): 23-37. Pal, Ranjan and Quizon, Jaime. Factor Costs, Income and Supply Shares in Indian Agriculture. AGR-ARU Discussion Paper 16. Washington, DC: World Bank, December 1983. Quizon, Jaime and Binswanger, Hans. "Income Distribution in India: The Impact of Policies and Growth in the Agricultural Sector". World Bank: Washington, DC, December 1983 (mimeographed). Scobie, Grant. Government Policy and Food Imports: The Case of Wheat in Egypt. Research Report No. 29. Washington, DC: International Food Policy Research Institute, 1981. Sen, Amartya. Poverty and Famines: An Essay on Entitlement and Deprivation. Oxford: Clarendon Press, 1981. Reutlinger. Shlomo and Selowsky, Marcelo. Malnutrition and Poverty: Magnitude and Options. World Bank Staff Occasional Papers 23. Baltimore: The Johns Hopkins University Press, 1976. DISCUSSION PAPERS AGR/Research Unit Report No.: ARU 1' Agricultural Mechanization: A Comparative Historical Perspective by Hans P. Binswanger, October 30, 1982. Report No.: ARU 2 The Acquisition of Information and the Adoption of New Technology by Gershon Feder and Roger Slade, September 1982. Report No.: ARU 3 Selecting Contact Farmers for Agricultural Extension: The Training and. Visit System in Haryana, India by Gershon Feder and Roger Slade, August 1982. Report No.: ARU 4 The Impact of Attitudes Toward Risk on Agricultural Decisions in Rural India. by Hans P. Binswanger, Dayanatha Jha, T. Balaramaiah and Donald A. Sillers May 1982. Report No.: ARU 5 Behavioral and Material Determinants of Production Relations in Agriculture by Hans P. Binswanger and Mark R. Rosenzweig, June 1982, Revised 10/5/83. Report No.: ARU 6 The Demand for Food and Foodgrain Quality in India by Hans P. Binswanger, Jaime B. Quizon and Gurushri Swamy, November 1982. Report No.: ARU 7 Policy Implications of Research on Energy Intake and Activity Levels with Reference to the Debate of the Energy Adequacy of Existing Diets in Development Countries by Shlomo Reutlinger, May 1983. Report No.: ARU 8 More Effective Aid to the World's Poor and Hungry: A Fresh Look at United States Public Law 480, Title II Food Aid by Shlomo Reutlinger, June 1983. ReDort No.: ARU 9 Factor Gains and Losses in the Indian Semi-Arid Tropics: A Didactic Approach to Modeling the Agricultural Sector by Jaime B. Quizon and Hans P. Binswanger, Sepcember 1983, Revised May 1984. Recort No.: ARU 10 The Distribution of Income in India's Northern Wheat Region by Jaime B. Quizon, Hans P. Binswanger and Devendra Gupta, Augusz 1993. Revised June 1984. Report No.: ARU 11 Populatton Density, Farming Intensity, Patterns of Labor-Use and Mechanization by Prabhu L. Pingali and Hans P. Binswanger, September 1983. Reuort No.: ARU 12 The Nutritional Impact of Food Aid: Criteria for the Selection of Cost-Effective Foods by Shlomo Reutlinger and Judit Katona-Apte, September 1983. -2- Discussion Papers (Cont'd.) Report No.: ARU 13 Project Food Aid and Equitable Growth: Income-Transfer Efficiency First! by Shlomo Reutlinger, August 1983. Report No.: ARU 14 Nutritional Impact of Agricultural Projects: A Conceptual Framework for Modifying the Design and Implementation of Projects by Shlomo Reutlinger, August 2, 1983. Report No.: ARU 15 Patterns of Agricultural Protection by Hans P. Binswanger and Pasquale L. Scandizzo, November 15, 1983. Report No.: ARU 16 Factor Costs, Income and Supply Shares in Indian Agriculture by Ranjan Pal and Jaime Quizon, December 1983. Report No.: ARU 17 Behavioral and Material Determinants of Production Relations in Land Abundant Tropical Agriculture by Hans P. Binswanger and John Mctntire, January 1984. Report No.: A.RU 18 The Relation Between Farm Size and Farm Productivity: The Role of Family Labor, Supervision and Credit Constraints* by Gershon Feder, December [983. Report No.: ARU 19 A Comparative Analysis of Some Aspects of the Training and Visit System of Agricultural Extension in India by Gershon Feder and Roger Slade, February 1984. Report No.: ARU 20 Distributional Consequences of Alternative Food Policies in India by Hans P. Binswanger and Jaime B. Quizon, August 31, 1984. Report No.: ARU 21 Income Distribution in India: The Impact of Policies and Growth in the Agricultural Sector, by Jaime B: Quizon and Hans P. Binswanger, November 1984. Report No.: ARU 22 'Population.Density and Agricultural Intensification: A Study of the Evolution of Technologies in Tropical Agriculture, by Prabhu L. Pingali and Hans P. Binswanger, October.17, 1984. Report No.: ARU 23 The Evolution of Farmining Systems and Agricultural Technology in Sub-Saharan Africa, by Hans P. Binswanger and Prabhu L. Pingali, October 1984. Report No.: ARU 24 Population Density and Farming Systems - The Changing Locus Of Innovations and Technical Change, by Prabhu L. Pingali and tans P. Binswanger, October 1984. Repor,t No.: ARUY 25 The Training and Visit Extension System: An Analysis of Operations and Effects, by G. Feder, R.H. Slade and A.K. Sundaram, November 1984. -3- Report No.: ARU 26 The Role of Public Policy in the Diffusion of New Agricultural Technology, by Gershon Feder and Roger Slade, October 1984. Report No.: ARU 27 Fertilizer Subsidies: A Review of Policy Issues with Special Emphasis - on Western Africa, by Haim Shalit and Hans P. Binswanger, November 1984. Report No.: ARU 28 From Land-Abundance to Land-Scarcity: The Effects of Population Growth on Production Relations in Agrarian Economies, by Mark R. Rosenzweig, Hans P. Binswanger, and John McIntire, November 1984. Report No.: ARU 29 The Impact of Rural Electrification and Infrastructure on Agricultural Changes in India, 1966-1980, by Douglas F. Barnes and Hans P. Binswanger, December 1984. Report No.:- ARU 30 Public Tractor Hire and Equipment Hire Schemes in Developing Countries (with Special Emphasis on Africa). A study prepared by the Overseas Division, National Institute of Agricultural Engineering (OD/NIAE), by P.J. Seager and R.S. Fieldson, November 1984. Report No.: ARU 31 Evaluating Research System Performance and Targeting Research in Land Abundant Areas of Sub-Saharan Africa, by Hans P. Binswanger, January 1985.