The distribution of income in India's Northern wheat region

DISCUSSION PAPER REVISED Report No.: ARU 10 The Distribution of Income in India's Northern Wheat Region by Jaime B. Quizon, Hans P. Binswanger and Devendra B. Gupta Research Unit Agriculture and Rural Development Department Operational Policy Staff World Bank June 1984 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 consultants 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. The Distribution of Income in India's Northern Wheat Region In an earlier paper, we had developed a unified approach to income distribution questions in agriculture, explored its properties theoretically and sketched how such a model could be empirically imple- mented and extended (Quizon and Binswanger, 1983). In the theoretical enquiry we showed how crucially dependent distributional outcomes of policies, programs and technical changes are on the final demand condi- tions - elastic or inelastic demand - on factor supply conditions and on factor mobility assumptions. The theoretical model was very simple and contained only one agricultural commodity and two or three factors. Furthermore, it con- centrated largely on output and factor price effects, ignoring secondary income feedback loops. Because the present paper is a numerical imple- mentation of the unified approach, we are now able to overcome the limitations. We present a general equilibrium model of the agricultural sector of the northern wheat region of India. The model contains four agricultural outputs with sharply different demand characteristics. It deals explicitly with the price and quantity of land, labor, fertilizer and draft power. All output and input prices are endogenously determined, as are the corresponding quantities (except for land). In the rural area, we distinguish four income groups, from landless to large farmers. The urban consumers of northern India are treated as a single group, as are consumers in the rest of India. -2- Real incomes of each of these groups are derived by summing up their factor incomes and deflating it by an endogenous income-group-specific price deflator. These real incomes then determine food demand'of each group. We therefore feed back the real income effect of all endogenous factor and food price changes into the final agricultural demand. In section 1, we present the structure of the model. Section 2 describes the origin of all parameters used for the model. The large majority of them were econmetrically estimated to be consistent with the model structure. In section 3, we present results of subjecting the model to a variety of exogenous shocks and policies. These include demographic and urbanization scenarios, technical changes, agricultural investment programs, weather changes, food aids and supply changes in the rest of India, and a variety of taxation and inc-ome distribution scenarios. We analyze such a large number of scenarios primarily to illustrate the model's capability and flexibility. The present paper is a preliminary exercise for two reasons. First, we present results which refer exclusively to the Northern Wheat (henceforth, NW) region of India, including its urban population. Two versions of the model are analyzed. In version A, the NW region is treated as if it were a completely closed economy, i.e., whatever output changes occur have to be consumed by the rural and urban population of the region. Inevitably, therefore, food demand faced by producers is relatively inelastic and supply shocks will result in relatively large price changes. In version B of the model, the NW region is assumed to supply the all India market for agricultural outputs. Demand of rural and urban consumers in the rest of India (henceforth, ROI) responds to price changes. Producers in the NW region, therefore, face much more elastic consumer demand. However, agricultural supply from the ROI region is still treated as exogenously given. Furthermore, India as a whole is treated -3- as closed with respect to international markets. We are in the process of building a four region model of India within which each region's supply is treated endogenously and where trade with the rest of the world will be -explicitly modeled. Second, while the parameters used in the model come from careful econometric inquiries, the weights and shares of income classes in factor supplies and factor demand have been pieced together from a variety of sources. We have nearly completed a reestimation of these shares and weights using the 1971 NCAER/ARIS survey within which we have divided India's rural population into four per capita income quartiles. The results presented here therefore have a tentative character. Apart from showing a number of implications of a closed economy scenario, i.e. one which aims at self sufficiency, they are designed to illustrate the types of issues which can be analyzed with this class of models. -4- THE MODEL The model described herein is a one region model of the Northern wheat-region of India, comprising the wheat growing states of Punjab, Haryana, Himachal Pradesh, Jammu and Kashmir and western districts of Uttar Pradesh 1/. Its immediate predecessor is a similar one-period model for the Semi-Arid Tropical (SAT)region of India reported in Quizon and Binswanger (1983b). The model described there differs from the current one in that distinctions between classes of producers are now introduced. Rural producers are distinguished according to their operational land holdings. All urban consumers are treated as a separate group. ^hese agricultural groups have different holdings of agricultural factors of production. Their agricultural incomes depend on the size and the factor payments associated with use of these holdings. Further, consumption patterns differ across groups because their real incomes differ. Although all producer groups are assumed to face the same markets where aggregate demands are equated to aggregate supplies, the producer group-specific effects of any exogenous shock in agriculture would differ across groups because of differences in the sources of income and in the patterns of consumption among them. The current model thus allows an explicit treatment of equity issues. The model is an extension of the unified approach 1/ We are also formulating a model for Indian agriculture with four agroclimatological regions, namely the Northern Wheat, the Eastern Rice, the Coastal Rice and the Semi-Arid Tropical regions. Appendix Table 1 lists the geographical divisions of these regions. - 5-- described in Quizon and Binswanger (1983a). The key feature of the model is that prices and quantities of agricultural outputs are endogenous, but the model differs from a economy wide general equilibrium model in that the nonagricultural sectors are not considered. Producer behavior is represented by a system of output supply and factor demand equations called the producer core. Analytically the producer core is derived from a variable profit function R* = H*(V, Z, T), where H* is maximized variable profits, V=(P,W) is the vector of prices of outputs (P) and variable inputs (W), Z is a vector of fixed inputs and T is a technology index. The output supDlv and factor demand curves are derived from 7* via Shepards lemma i.e. the vector of outputs (Y) and (negative) variable ijputs (-X) is written as Q = [Y,-X] = . In terms of rates of changes they are written as 3V (1.1) Q, = E. .V' + E3. Z* + E' iEO, VI S . j .j 3 -g 8 i 0 is the set of outputs and VI the set of variable inputs. The prime notation as a variable X' indicates the total rate of change over time of variable X. The star notation X* refers to the rate of change of an exogenous variable or the exogenous component of an endogenous variable. a.. are the elasticities of supply/demand of an output/factor i with respect to a price j. The Zs are exogenous variables and fixed inputs affecting producer behavior and the a. are elasticities with respect to those fixed inputs. Some of the Z variables are subject to government policy. Ei 3Q t are the technology shifters of the supply/factor demand i 3T Q.i 3t . 1* equations, holding fixed inputs constant (for a detailed discussion of these technology concepts used see Binswanger (1983)). -6- Output demand is treated in a more disaggregated fashion. Let k=1, . . ., K refer to income groups. Then, total final demand is defined as (1.2) Y E Yik icF0 k where Y.i is the total demand of consumer group k. Rewriting (1.2) in terms of changes (1.3) Y= i E. Yik i0 i 'ik YIk k where X. = Y. /Y. is the proportion of commodity i consumed by income ik ik i group k. The consumption of each income group is described by an income group specific consumer demand system (1.4) Y = Nk 4(,m,k) where the underbars denote a column vector of the variable, e.g., k 1 2 O Nk is the population in income group k and y ik. is the per capita demand which depends on out?ut prices and per capita income of the income group. Transforming each equation of (1.4) into rates of changes leads to (1.5) Y' = y' + N' = aa P! + a. m' + y* + N' k k k 'ak imkT ik k i,jEO yk is an exogenous change in per capita demand of income group k, and Ik the a.. and a. are the price and income elasticities of final demand. We assume that the population in each income group grows at an exogenous rate N*. But the rural population grows via immigration or via diminished emigration, a process which we assume to be respcnsive to the real wage rate w prevailing in the region. Differentiating Nk with respect to time and the real wage and converting to rates of changes, we obtain (1.6) N' = £ w' + N* = (W' - T') + N* k mk k mk.L k k - 7 - where emk is the migration elasticity-into (or for the urban group, out of) the specific income group with respect to the real rural wage. And Pk is 1/ an income group specific price deflator defined below - . The migration elasticities are discussed in Appendix 1. Let the rural income groups be indexed k=l .. . . K-1 and the urban group by'K. Total labor supply K-1 to agriculture is L = E Lk, or in rates of changes k=1 K-1 (1.7) L' = X L' klLk k k=1 . where XLk L k/L is the proportion of labor supplied to agriculture by income group k. Labor supply of income group k is Lk = N where Zk 0k k k k is total labor supply per person (the product of the labor participation rate and the effort per participant). Differentiating with respect to the real wage and time and converting to rates of changes, we find (1.8) L' = E Zkv + * + N' = E - P') + t* + N' k k k Zk L k k where EZk is the total labor supply elasticities of income group k and Z* is an exogenous shi-fter in the labor supply to agriculture of income group k. The supply of bullocks is similarly aggregated K-1 as X. = E X . Rates of change are aggregated as usual 1ik k=l K-1 (1.9) Xi X X i = Bullo.ks k=1 i 1/ If the urban region or the donor region also experiences the same price changes, then (1.6) should be in terms of nominal wages only. The supply of each input is only dependent on its own price, W., therefore 1 (1.10) Xi Ei i k) +i = Bullocks While the model contains many Z variables, such as irrigation, rainfall etc., we treat land (indexed as the first fixed factor Z ) as, the only fixed factor which is a recipient of rent. The change in the rental rate of land (or residual profits) S' is derived residually from the profit function, a derivation given in detail in Quizon and Binswanger (1983b). (1.11) S' = E6.V! + E6.Q, - Zt iEO, VI ,.11 1 1 1 1 0.V. 1 1 where = are variable profit shares, which are positive for outputs and negative for inputs. Changes in income group specific consumer price levels P' can k be related to the endogenous changes in agricultural output prices as follows (1.12) P' =E P' + z P* k ik i NAk NA where pik is the share of total consumer expenditures spent on commodity i by income group k. The subscript NA refers to nonagricultural commodities. The GDP deflator P is derived inthe same way by dropping the k subscripts. Nominal per capita income of income group k is Mk and is defined as the sum of all net factor incomes accruing to the group plus nonagricultural incomes MN (1.13) Mk E X ik Wi + zIkS + MNk iEVI 1/ We treat all rural labor supply as an "agricultural" income here, because we assume wage equalization between the agricultural and nonagricultural rural labor market. -9- Real per capita incomeis derived by dividing by the number of people and the consumer price index i.e. (1.14) mk =Mk/TkNk Differentiating (1.13) and (1.14) totally and converting to rates of changes leads to (1.15) ' Z=k E 6. (Wi + X! ) + 6 (S' + Z* ) - P' - N' + 6 MN* k Ik i 1k Sk 1k k k Mkk where the 6. are the shares of net income arising from the respective I source and Z*k is the exogenous rate of growth of land supplied by group k. Real per capita income of North India's rural and urban population is defined as (1.16) m = ANk 'k k where XNk = Nk/ENk are the initial shares of group k in the total population. k Differentiating and converting into rates of changes leads to (1.17) m' = XNk- mk k= k k k where vk is the proportion of real income accruing to group k. Note that m' is not equal to the conventional definition of a char-ge in real per capita income"which would be (1.18) fi' = M' - N' - P' where P' is computed from the equivalent of(1.12) but dropping the k subscript, and M is defined as in (L.13 ) but again dropping the k subscript. The difference between (L.1) and (.18) is that L11) utilizes real income weights, where each group's real income is deflated by a group-specific price deflator. (.17) is thus much closer to a measure of a change in real per capita welfare than L.18 ). - 10 - Version A of the model treats the Northern Wheat region as a closed economy. The full model consists of equations (1.1), (1.3), (1.5), (1.6), (1.7), (1.8), (1.9), (1.10), (1.11), (1.12), (1.15), and either (1.17) or (1.8). The equation system can be exhibited in matrix form (1.19) GU' = K* where G is a square matrix of elasticities and shares, U' is the column vector of endogenous variables and K* is a column vector of exogenous shifter variables. (For simpler examples of such full systems see Quizon and Binswanger (1983a and b)). The effect of a shift in an exogenous variable on the endogenous variables in the system can be solved as (1.20)* U' = G 1K* which exists so long as the matrix G is nonsingular. In Table 1 we exhibit a sample case of version A of the model in matrix notation. It contains only one output, two variable factors of production and two income groups. In version B of the model, agricultural producers of NW region supply the all India market. The subset of output supply equation (1.1) is changed to include an exogenously given supply from producers in the ROI region. On the demand side, we add another group of consumers, the ROI consumers in equation (1.2) which are also represented by their own consumer demand system as in equation (1.3). Evidently, the X weights of the ik consumer groups of the NW region are much smaller in version B than in version A. Note, however, that in the results, we are concerned only with the national income of the NW region, i.e. in equations (1.13) to (1.18), we do not include the incomes of the ROI residents. - 11 - DATA The data used to compile a G matrix for the Northern Wheat region of India come from a variety of sources. Only four agricultural outputs, namely, rice, wheat, inferior cereals and othe crops are separately identified. The agricultural commodities are exhaustive in that they account for all crop production in the agricultural sector. Livestock products are aggregated with all other commodities into "nonagricultural commodities". Five producer household groups were distinguished based on operational land holdings. These household groups are (a) marginal farmers or households with less than one acre of operated land, (b) small farmers or households with one to less than five acres of operated land, (c) medium farmers or households with five to fifteen acres of operated land, (d) large farmers or households with more than fifteen acres of operated land, and (e) urban (non-rural) households. The first four groups refer to rural - 12 - households only and are exhaustive of the rural population. Our categories of producer groups are not only quite common to the empirical literature on Indian agriculture, they also essentially capture prevailing differences in the ownership of producer inputs and in patterns of agricultural consumption across households. The following factor inputs to crop production were identified: labor, bullocks, fertilizer and land. Land is treated as a fixed factor of production. Fertilizer is the only input for which supply is not income group specific. Farm machinery and implements are not explicitly introduced as inputs in the model, although their input shares were taken into account when various other input shares were computed. The commodity specific output supply and the fertilizer and labor demand elasticities are from Evenson's (1981) study of North India districts. His original estimates were derived after he imposed all regularity constraints on the system of supply-demand equations as obtained from a twice differentiable variable profit functic.., except for the condition of convexity of the resultant Hessian matrices. Evenson's estimates were therefore adjusted in an ex post manner in order to satisfy this convexity constraint, following trial and error procedures described in Quizon and Binswanger (1982b). Appendix Table II shows Evenson"s elasticity estimates before and after adjustments for the convexity constraint were made. The bullock power demand elasticities have been estimated in Evenson and Binswanger (1981). Only the own price elasticity and the cross price elasticity with respect to labor are available for bullock power demand. - 13 - The output demand elasticities are from Binswanger, Quizon and Swamy (1982) and are averages for all India. Original price coefficient estimates in the reported demand equations in this study were first adjusted following above mentioned procedures to satisfy convexity restrictions. Expenditure group specific output demand elasticities were then computed using these adjusted convex price coefficient estimates. The eight different expenditure groups identified in the afore- mentioned study are not the same as the producer groups desired for use in this paper. Therefore, a mapping of the producer groups onto the expenditure groups had to be done. Real per capita producer group specific incomes were matched with similarly computed incomes for the expenditure groups. Matchings were not one to one. Rather, the per capita income of each of the five producer groups was taken as the weighted average of the two expenditure groups with the closest reported per capita incomes. The weights computed in this matching were subsequently used to obtain producer group specific elasticities from the expenditure group specific elasticities, i.e., the producer group elasticities were computed as weighted sums of the expenditure group elasticities. Real per capita income for the urban group was obtained from the 28th Round of the National Sample Survey (28th NSS), Table in Consumer Expenditures, while incomes for Ehe rural producer group were computed from the 28th NSS and the per capita income figures by producer group from the 1970-71 National Council for Applied Economic Research (NCAER) Additional Rural Income Survey 1/ 1/ Both urban and rural real per capita incomes are from the 28th NSS. The NCAER-ARIS data was used to compute rural producer group specific per capita incomes given the average rural real per capita incomes from the 28th NSS. All computed incomes refer to what we previously defined as the Northern Wheat Region. - 14 - The Xik output consumption weights in (1.3) are from the 28th NSS. This data set gives the number of persons per household and the mean per capita value of consumption of good i by urban versus rural and also by expenditure group for the-NW states. The number of persons per household and the per capita consumption of good i for each of the rural producer groups were computed using the same income based weights employed to map producer groups onto expenditure groups. Total average household consumption by commodity and by producer group was then computed. Given the percentage distribution of households between rural and urban (from the 28th NSS) and the percentage distribution of rural households across rural producer groups (from the NCAER-ARIS), the shares of each producer group in the total consumption of good i was computed. The Xik input supply weights in (1.7) and (1.9) are basically from the 26th Round of the National Sample Survey (26th NSS), Tables on Landhold- ings, All India. From this data set (a) the amount of land owned, (b) the value owned capital (from machinery and equipment and non-draft animals), (c) the number of working buffaloes and cattle owned, (d) the number of house- holds operating and not operating land by rural and urban and by rural pro- ducer group can be immdeiately computed. The average household ownership of (a) to (c) above by the same breakdowns can likewise be easily obtained. Hence, given the North India distribution of households across producer groups earlier computed, the distribution of ownership of land, capital and draft animals by producer group (for the NW region) was obtained. The urban versus rural distribution of agricultural labor was assumed to be equal to the distribution of land operating households between urban and rural. The rural share in this distribution was further allocated across rural producer groups using both the size and the percentage distribution of - 15 - households across these groups as weights. The parameter i in (1.11) is actually the ratio of two shares (see Quizon and Binswanger (1983b)). Let si denote the value share of input i in the total cost of production or the value share of output i in total Si revenue. Then, if s is the share of land in cost, 4 is equal to --. In this z i sz Z study, the values of s., for all iEVI, and of s were assumed to be equal to the share of factor i in total crop income computed to be as followsi/: land, 29.28 percent; labor, 42.33 percent; capital, 16.02 percent; and draft animal labor, 12.37 percent. The output shares si, iEO, were computed from the output prices and quantities used by Evenson (1981) to estimate North India output supply elasticities. The t ik weights in (1.12) are computed from the estimated total consumption by commodity and by group, already solved for in getting the x. of (1.3). ik The 6 weights of (1.15) are computed by first allocating total ik income (assumed equal to one) to the different factors of production (land, labor, capital and draft animals) using previously computed shares in crop income by factor of production. This was then reallocated across all groups using the distribution of each factor input supplied by each producer group, i.e., the Aik computed for (1.7) and (1.9), as weights. The resulting matrix therefore, showed the distribution of total crop income by producer group and 1/ This distribution of crop income by factor of production uses farm management data for Haryana in Handal and Grover (1976). - 16 - by factor of production. Given the proportion of crop income in total income in each producer group (from the NCAER-ARIS- ), the share of income from input i in producer group k (d ik) is solved straightforwardly. The proportion of the 'population in group k, XNk, in (1.16) was computed from earlier computations used to solve for Xik in (1.3). Finally the weight vk of (1.17) was likewise obtained from earlier computations of real per capita incomes by producer group and of the distribution of the population across the same groups. The only unaccounted parameters in our G matrix thus far are the input supply elasticities, i.e., E ik in (1.6), (1.8) and (1.10). Similar to Quizon and Binswanger (1983b), we assume Eik to be equal to 0.3 based on Rosenzweig's (1980) econometric estimates. The migration elasticities, Emk, are computed from Dhar (1980). Appendix I reports these elasticities and also details how this was done. For bullock labor, the own price elasticity is assumed to be equal to 0.2882, i.e., the value weighted sum of the own price elasticities of supply for agricultural outputs. This follows from the notion that bullocks are reproducible out of agricultural output. Finally, the fertilizer supply elasticity is set at 4.0, a high value which reflects opportunities for international trade. All important elements- of our G matrix are given in Appendix Tables 2 to 8. The G matrix can be reconstructed from the submatrices shown in these tables by following the matrix form for G given in Table 1. 1/ In the urban group, this proportion is from summary tables of a 1975 NCAER survey, similar to the 1971 NCAER-ARIS, but which investigates urban incomes. - 17 - The S matrix (equation (1)) for the shifter variables are ig from Evenson (1981) and are listed in Appendix Table 9. The full K* matrix. of exogenous shifter variables (equation (1.19)), i.e., the rightmost column of Table 1, can be reconstructed from this and other previous Appendix tables. - 18 - RESULTS In Tables 2 to 7, we provide the growth and distributional effects from a few simple simulation exercises. In designing these simulations, we had a time frame of one decade in mind. For example, in simulation 1.1, both population and labor force growth are reduced by 10%. This corresponds to the effects over a decade of a slowdown of the population and the labor force growth rate of a little less than 1% per year. The effects shown are the cumulative changes over a decade which would result in the growth of the endogenous variables as a result of this slowdown in population and labor force growth. These changes are with respect to the ceteris paribus development paths of the endogenous variables. We do not know or model these paths at the present time. Demographic and Urban Growth Scenarios In demographic scenario 1.1, population growth in North India (rural and urban) is reduced by 10% (over a decade). This is accompanied by a backward shift in the rural labor supply of 10%. Nominal urban income is reduced by 10% as well, i.e., the exogenous components of nominal urban per capita income are left constant. In the first row of Table 2, we see that real per capita income growth (in the NW region) would accelerate by nearly the same amount i.e., about 7.5%, regardless of whether the region is closed (version A) or whether it supplies the all India market( version B). The major difference in the two cases have to do with how this growth is partitioned into quantity and price effects. In version A, where final demand is inelastic, aggregate -19- agricultural output and the GNP deflator both decline, by 2.51% and 3.70% respectively. The increased scarcity of labor leads to a sharp rise in real wages (13%). Labor employment declines by about 5%. This is only half the decline in the exogenous component of labor force growth. The reduced demand and lower aggregate prices of agricultural output lead to a sharp decline in the real residual profits of 28%. The output and factor price effects largely drive the distributional outcome. Rural landless households experi- ence the major real income gains of nearly 15% whereas large farmers lose a little less than 2%. The urban households of North India gain by 7.5% because nominal per capita urban incomes are initially held constant and real gains arise from the decline in the urban price deflator. The popula- tion in each of the agricultural groups declines by about 8.6% (not shown) while that of the urban group declines by 15%, i.e., improvements in the real rural wage slow down the rural to urban migration process within North India considerably. Per capita cereal consumption- of all groups, except the large farmers, increase, with the gains somewhat smaller than the corresponding income gains. In version B, final demand faced by producers of the NW region is far more elastic, and the per capita income effects of lower population growth tend to dominate the price effects of lower numbers of consumers. The GNP deflator increases by a minimal amount which is the combined effect of increased prices for rice and wheat (the relatively high income elasticity crops) and a 1/ Per capita cereal consumption is the closest we can now come to measure changes in per capita calorie intake. At a later stage, we intend to convert the output data into calories. - 20 - reduced prices for coarse cereals, and "other crops". The real wage rate increases a little more than in version A but the decline of land rents is now only 20% rather than 28%.. The net effect is higher incomes for large farmers, instead of the slight erosion in version A. Urban consumers gain less, as they experience a slight increase in food prices. Finally, ROI consumers do not gain at all, as the GNP deflator rises slightly, i.e., all the gains are captured by the NW region, regardless of whether it does or does not supply the rest of the country. Simulation 1.2 is an urbanization scenario. Rural population in the NW region is assumed to decline by 10% while urban population increases by 30.11%, enough to absorb the rural population. Nominal urban income is increased by 30.11%, in order to initially hold nominal per capita income constant, otherwise the scenario would be unrealistic. The consequences of reducing the number of producers while leaving the number of consumers constant differ sharply depending on whether NW con- sumers have to give their own reduced number of producers the incentive to maintain the aggregate output nearly constant (version A) or whether the NW region can compete for agricultural supplies with ROI consumers (version B). In version A, agricultural prices have, on balance, to rise sharply. They drive the GNP deflator up by 28%. Only coarse cereal prices decline because increased per capita incomes drive down coarse cereal demand. In version B, price increases are rather small and the GNP deflator rises by only 2%. The reduced supply from NW producers is made up by squeezing ROI consumption. In North India, the reduction in agricultural population leads t, sharp real.wage increases in both versigns A and B (15.7% and 12.6% respectively). But the sharp increase in prices in version A enables residual profits to rise sharply as well (+43%). In version B, on the other hand, farmers have to pay - 21 - higher real wages but do not receive corresponding output price increases. Land rents or residual farm profits therefore decline. The differential food price and land rent effects determine the differences in income distribution effects. In version A, large farmers gain by nearly 24%, urban consumers lose by 15.5%, while the other groups fall in between. In version B, the landless and small farmers are the principal gainers. Urban households also become net gainers. The urben to rural migration induced by the rise in the rural wage rate reduces the increase in the number of urban residents below what the initial exogenous shift implied. This effect dominates the negative impact of rising food prices. 4 Simulation 1.3 combines the two previous scenarios. Overall pop- ulation growth rate declines by 10%, but the decline is accompanied by a rural to urban migration so that the rural population is initially stabilized while the urban population now initially increases by 20%. Evidently the outcomes of the. e shocks are a combination of the two previous simulations. A large overall gain in real income results (+12.7% for version A and +14.4% for version B). Version A and B again differ primarily in their price effects. Sharp increases in agricultural prices coupled with a modest increase in land rents are the results in version A. In version B, prices stay dlose to con- stant and land rents drop sharply (-34%). In version A, therefore, all agri- cultural groups experience real income gain of about 20%, while the urban group loses out. In version B, on the other hand, net labor suppliers (the landless and small farmers) capture the largest gains while large farmers gain little. The net income gain of urban groups is 8.7%. Simulation 2.1 lets the exogenous component of urban income increase by 10%. In version A, aggregate agricultural output increases only slightly (+.56%) due mainly-to the substantial increase in the aggregate price level (+6.99%). In version B,,the urban NW population satisfies its increased demand largely - 22 - from the rest of India. Since in both versions quantities produced in NW rise little, real wages are largely unaffected. But in version A the real land rent rises by 15.3%, while it stays almost.constant in version B. In version A, landless households lose 2.2% of their real income while large farmers gain about 5.6%. The ilrban group has to share its initial income gain of 10% with the large farmers. The initial urban gain of 10% is reduced to a real gain of about 4.3% because of the agricultural price increases. In version B, on the other hand, all real rural incomes stay about constant and the urban group retains nearly all (9.2%) of its initial income gain. Scenario 2.2 combines the three individual scenarios: slowdown in populazion growth, faster urbanization and urban income growth. The effects are largely additive. Under version A, real urban incomes still decline by 3.7% while incomes rise by about 22% for all rural groups. In version B, urban groups gain nearly as much as the landless and small farmers, while large farmers lag behind. An important overall feature of all the scenarios involving re- duced population growth and increased urbanization (except scenario 2.1) is that the poorest group, the landless farmers, always gains regardless of the inter-regional trade assumption. Technical Change Scenarios In simulations 4.1 to 4.4 of Table 3, yields of individual crops or crop groups rise by 20%, a change --orresponding to a major varietal shift. In scenario 4.5 the yield gain is smaller, only 10%, but is distributed evenly across all crops. We start by discussing this last scenario first. The contrast between the closed version A and version B is again striking. Technical change makes possible increased agricultural output (+9.6% in version A and +11.1% in version B). When the extra supply has to be consumed - 23 - locally (version A), all agricultural prices drop sharply (-15.4% to -41.5%) resulting in a reduction of the GNP deflator of about 20.6%. Therefore real per capita income rises more sharply (+7.97%) than in version B (+6.2%) where price effects are quite modest. Real wage rates are not very much affected in either case, but the real residual profit is. It drops by 10.6% in version A and rises by nearly 29% in version B. When prices do not decline, the extra quantities made possible by the te -hnical change translate into- sharp gains for the landowners. On the other hand, when the economy is closed, the negative price effects dominate the positive impact on land rents that are due to the output effect. Distributional outcomes are again largely driven by the differences in price effects. In version A, the largest gainers are urban consumers (+16%) followed by the landless (+9%). Large farmers' incomes are unaffected. In version B, the income distribution impact is exactly the reverse. Urban consumers and the landless gain only 2.2% and 3% respectively, while large farmers gain 14%. The cereal consumption effects are also larger in the closed version A as both the relative price and income effects lead to substantially increased consumption of cereals. In version B, the small price effects have only a minor impact. Yield increases in individual crops (or crop groups) of 20% result in real income gains whose magnitude largely reflect the share of each commcdity in NW agricultural output. Overall income gains are largest for the aggregate "other crops" (8.8% in version A and 7.1% in version B) and lowest (less than one percent in versions A and B) for coarse cereals. The distribution of these gains, on the other hand, is primarily a reflection of the demand char- acteristics of the crops. As shown in Appendiz Table 3, the income elasticities for coarse cereals are negative. Income elasticities for rice range from about 0.4 to 0.7, those for wheat are close to one and those for "other crops" -24- slightly exceed one. Price elasticities for coarse cereals are the most in- elastic (-0.01 to -0.5). The other agricultural commodities have more elastic demands with price elasticities ranging from -0.72 to -1.02. If we consider version A only, these differences sharply affect the distributional outcomes. A rise in output of a commodity with high income and price elasticities leads to a smaller price decline (and corresponding agricultural profits decline) than a. rise in output in an inelastically demanded crop. Yield gains in rice reduce the rice price by 78% and lead to a massive diversion of resources from wheat to rice. Wheat production declines by some 6.9% while rice production increases by about 28%, i.e., rice absorbs nearly all resources released by wheat, some released by other crops. Per capita cereal consumption increases by nearly 9%. As final demand is in- elastic, employment and real wages decrease, but only minimally compared to the sharp drop in the real residual farm profits. Urban consumers are the gainers. Their income rises by 17.8%, while that of large farmers declines by 16.7%. The landless gain nearly 9% while small farmers manage to stay about even. Cereal consumption increases in all income groups, i.e., the price effects dominate the income effects even for the losing large farmer. Although in version A the real income effect of technical change in wheat is the same as for rice, the relative price effects among crops are less sharp. Real residual profits therefore are less affected and decline by only 8.8%. Since relative prices of inputs and outputs change less sharply, income distribution effects are also less dramatic. While urban consumers gain only about 7.8% in real income, large farmers lose only 1%. For technical change in "other crops", the large income gains and the high income elasticity .of demand for this aggregate dominate the effects of a fairly low own supply price elasticity. Unlike in scenario A for wheat and - 25 - rice therefore, residual profits increas by 23.8%. The distributional impact is the reverse of the rice and wheat scenarios. Large farmers become the major gainers (+13.2%), followed by urban consumers (+9.2%). Landless households gain by 5.3% only. Coarse cereals have a negative income elasticity that becomes more negative with higher incomes. They also have the lowest own price elasticity of demand among all crops. Under version A, technical change in coarse cereals, therefore, has quite unanticipated effects. It leads to a very large price drop for coarse cereals (-6.3%), but only a minimal in- crease in its output (+3.8%). This large relative price drop leads to income gains which in turn drives up the price of wheat and rice by 22.3% and 24.4% respectively. These price effects lead to an increase in the land rent of 11.1% and to somewhat counterintuitive income distribtuion effects. Urban consumers lose about 2.8%. For the poorest landless group, the gains from the drop in coarse cereal prices together with the slight gain in the wage bill of 0.86% result in a -modest gain of 0.25% of real income but a substantially bigger gain in cereal consumption of 1.7%. The biggest gainers are large farmers whose incomes rise by 4.1%. Compared to version A, price effects are much smaller in version B, for now very familiar reasons. Furthermore, residual farm profits rise in all cases as all aggregate demand curves are now price elastic. Large -farmers are inevitably the largest gainers with small farmers, the landless and urban groups, by and large, experiencing small income changes. In is interesting to note that real rural wage effects are always relatively small.in magnitudes, regardless of the nature of technical change or of whether we are in version A or B of the model. Under version A, tech- nical change in rice, wheat and "other crops" results in large price effects, and wage earners gain primarily as consumers. These price effects lead to - 26 - relatively large increases in cereal demands both via the price and real income terms of the Slutzky equation. In version B, workers' real incomes rise much less because they do not gain as consumers. Moreover, cereal con- sumption increases much less as they are not increased via price effects. When technical change is concentrated on "other crops", the price effects favor those crops at the expense of wheat and rice. Thus under sce- nario A, cereal consumption of all rural groups drops, despite their gain in real income. This however does not necessarily imply a reduction in nutri- tional standards as "other crops" supply calories and other nutrients as well. . Investment Scenarios In scenario 3.1 of Table 4, irrigated area in the NW region is assumed to increase by 10% because of investments in irrigation. Consider first the version B. Here, the output effects are very substantial; aEgregate output increases by 8.09%. The price effects are relatively modest (GNP deflator declines by 1.38%) and are concentrated in coarse cereals and wheat. The real wage and employment effects, 5.4% and 2.21% respectively, show that irrigation investment is labor using. In version A, sharper price drops follow from the increases in output made possible by irrigation. Real residual profits decline by 20% and the distributional outcomes follow from these. In both versions A and B, the landless always gain. Large farmers lose in version A but gain in version B. NW urban households gain much more in version A where price drops are sharper. The real per capita income gains from irrigation investments are larger in version A (5.7%) than in version B (4.7%). In the latter case, the ROI region-shares in the benefits of the modest price declines. In version A, all the benefits from the (sharp) price declines accrue only to the NW region. - 27 - Scenario 3.2 focuses on expanding capital inputs such as machines, tractors and livestock - via expanded rural credit for example - and on improv- ing marketing infrastructure. Both of these are accelerated by 10%. Real per capita income increases by about 4% in version A. In version B, the real per capita income gains for NW (+3.8%) are again smaller than in version A because the modest price declines transmit a part of the gains to the RO. In versions A and B, however, we see that these investments have virtually no impact on labor use and real wages. Income distribution effects here are also different from that arising from irrigation investments. They are more regressive than progressive because price effects are less negative and effects on real residual profits more positive when compared to similar effects arising from irrigation investments. Scenario 3.3 combines the two previous scenarios, but with a con- centration on irrigation which is accelerated twice as much (+10%) as capital and marketing (+5%). Since the direction and magnitudes of distributional . effects of scenarios 3.1 and 3.2 are different, their combined effects result in a progressive income redistribution for version A and a regressive income redistribution for version B. Overall, however, real per capita incomes in- crease by at least 6.6% in both cases. In Table 5, we combine the effects of the population scenarios with those of the technical change and investment scenarios. The question answered in the first column and the first row of Table 5 is the following: By how much (in percentage terms) should irrigated area have to increase in order to maintain real per capita income in NW at a constant level in the face of an.increase in the population growth of 10% (or more precisely an increased rate of population growth of about one percent over a decade)? - 28 - We can see that irrigated area will have to increase by 13.33% under the closed version A. Because ROI shares in the irrigation benefits under version B, the irrigation increase would have to be much larger (+15.35%) to maintain real per capita incomes in NW. From the first row, we see that both the required investments and the rates of single-crop technical changes required are very large. Only if all yields improve simultaneously are the required technical changes in a reasonable range (9.5% and 11.6% for versions A and B respectively). The policy goal, may not, however be to stabilize per capita in- come, but may focus instead on the poorest group, the landless. On the second line we see that, because irrigation has a labor using bias and because the technical changes initially reduce the demands for all inputs and are moreover not particularly labor using, it is generally more difficult to maintain the real income of the landless via technical changes than via irri- gation investments. A third goal may be the maintenance of real per capita cereal con- sumption either in the aggregate (line three) or for the most vulnerable group, the landless (line four). Note first that neither goal can be achieved by technical change in other crops, which is not surprising as this group competes directly for factors of production with the cereals. Second, it is very much easier to stabilize cereal consumption when price effects reinforce income effects, i.e., when the NW region is closed as in version A. Required investments and technical changes are often less than half as large as that needed to achieve the same goals in version B. In the last line of Table 5, we show what would be required to main- tain the real wage rate (this goal, however, may not be as important as maintain- - 29 - ing the real income'of the landless). This goal cannot be achieved at all with capital and marketing investments. Also, the technical changes required are either exorbitant or completely infeasible. In the face of population growth, it is hard to stabilize real wages even with a labor using investment such as irrigation. Scenarios with Rainfall, Food Aid and Increased ROI Production Scenario 5.1 of Table 7 traces the effect of a 20% shortfall in rainfall. It is therefore a short run scenario. The effects of rainfall on output have been estimated econometrically as part of the output-supply and factor demand system and are fed into the model just like any other supply shifter. The scenario traces what would occur if sowing rains were reduced and a certain amount of adjustment of production plans was still possible in the same year. The aggregate output effects are fairly modest, -0.24% in scenario A and -0.42% in scenario B, where part of the output shortfall can be made up via larger purchases from ROT. Rice output is most severely reduced in both versions. Because the food prices rise by about 3% in version A, real residual profits increase, which implies that farm households end up as net gainers while the landless and the urban groups lose. Their cereal consumption drops. In version B, the food price rises are quite small, and lard rents rise little. Qualitatively, the distributional impacts are similar to version A, but quantitatively they are much smaller. In scenario 5.2, NW wheat supply is shifted by 10%, simulating the effects of food aid. In version B, this implies an increase in national wheat supply of 5.62% as NW supplies about 56.2% of the national wheat market. When the NW region is closed, this results in a sharp drop in wheat and rice prices, and a reduction of 5.6% in NW wheat production. This reduction, however, is partly compensated by the increased rice and coarse cereal pro- duction associated with the relative price shifts and the income gains of - 30 - consumers. Aggregate agricultural output drops by only 1.3%. The distribution of the gains are very uneven. Urban consumers gain most while large farmers experience a 12.5% reduction in real income. The landless experience a net gain of 4.0%. In version B of the model, the distributional affects are qual- itatively similar, bt much smaller in magnitude. Moreover, all ROI consumers experience a real income gain. The increased incomes and the lower wheat and rice prices translate into substantial demand increases for both wheat and rice. Wheat output declines by only 2.1%, i.e., by only one-fifth of the amount provided by food aid. The output of all other crops increases so that the decline in aggregate output is now only 1.30%. In scenarios 9.1 to 9.4, which pertain only to version B, the output levels of one crop or crop group is exogenously increased by 10% in the ROT. In scenario 9.5, an increase of 5% is spread evenly among all crops. Since the income gain of ROI is exclusively a gain arising in consumption from price declines, ROI incomes increase in scenarios 9.1 to 9.5. In NW, where we also trace changes in producer incomes, this is not necessarily the case. Large farmers lose when supplies of rice, wheat, "other crops" and all crops increase. Landless laborers, on the other hand, gain sufficiently as consumers to be net gainers in all cases. The effects of supply increases in coarse cereals are rather inter- esting. NW is a rela-ively unimportant producer and consumer of coarse cereals, except for landless laborers. What drives the results are the initial income gains of coarse cereal consumers (mainly landless workers and the ROI). These are translated into increases in wheat and rice demand and corresponding price increases. These price changes in turn erode the gains of the landless and ROI, and turn NW urban groups into net losers (they consume very little coarse cereals). NW farmer groups end up as net gainers as they are mainly producers of wheat and rice, whose prices have risen. г 5 � 5 �Ii1�1!!! ! Vег�эlи� А оЕ ti�e Nп1е1 1п Ии[г1х Nu[a[lw� р� Ч' К' У' У1 У2 I.` 1.� 1.� N, Н; Кz b� и� и1 Se Pj Р2 � щ! ш' Р т' [Lcµ�c 5„'д,►у -�гУ -�[!. -�rы ' . Р� ��г� � + Еч 1нЬnг lkv.ud -'1.У • �L �.В � Ч� В't.�Lg + _L Lillайи lkw�ni -!� -� -9� 1 �� g�2g + д. Аудп:ьрСе U�mud 1 -�1 �2 У� V 4Уи�р l Ilалицl -а�,� 1 -l -п� У� У'� + q'iJA1�NA Сгшр '2 Uuuvцl -а�,2 1 . -1 -а� У2 2+ i3dA2'�ИА А�,"пЕи[м 1�Gor . 1 -�1 �2 L 0 GYu�p 1 lдYvr -[�� ! -1 [� . L� Z� . 4Yuip 2 1d1r�r -ь� 1 -1 �2 � �'1 Pwx,1.,t1u� 1 -�� -� и� � е - _и !'y.�lactui I - ш1 1 E�W Nj � � • ! PqxiLьtlr.n 1 -t� � е� !� � ' АцЬRьЧса tцllск� ' -�у! - 1ц ыа � 4Уа�р 1 ц�У Lгкkв -г�� ' Е1Эl , ы1 � 4Ушр 2 В,11одш -г� � � � Ы 2 l�э,д 1V:��г -�[ ,1. �i •r �. ,И 1 у' г+ Рг1т 1 -ьч� � Pi l4гц�ИА • PrIa:2 -�4t 1 PZ !�1'а� 1+�1 ш1 -11 l!l 1.l ' "_1ц ^�l 1 1 ш1 51�1 + 7tll'�1 . г�! т1 . -4г '4>г -�г � "gt -� ' ' °� '�z� + $�rf°"z 14:а1 ш "� -ц1 l т' U (iиw Р -�ц ' Р� uNA �НА (iыw. 1tr_и1 mim� -4 -дН -С� 1 -б� -Q� 1 1 щ' Q,'С'� + Д_� TABLE 2 DEMOGIIAPHIC SCLNARIOS SIMULATIONS SLOWER FASTER SLOWER POPULATION INCREASED COMBINATION POPUI.AFION URBANIZATION GROWTH AND FASTER URBAN OF 1.3 AND 2.1 GROWTH UHBANfZAI[ON INCOME 1.1 1.2 1.3 2.1 2.2 A a A B A 8 A a A El REAL NAT.PEA-CA41 INC.(N.W.) 7.696 7.253 5.101 7.133 12.699 14.386 2.1685 2.69557 14.866 17.082 TOTAL OUTPUT(N.W -2.507 -2.146 0.256 -1.703 -2.251 -3.B50 0.5601 0.05565 -1.891 -3.794 QUANTITIES OF : RICE PRODUCED -2 411 -2 902 -7.168 -3,941 -9,579 -6.743 -1.0645 -0.26823 10,644 -7.012 WHEAT PRODUCED -3 4BI -2.855 3.159 -1.848 -0.322 -4.703 1.3315 0.15693 1,010 -4.546 C CEREALS PRODUCED -11.570 -B.016 -22.155 -1.957 33.726 -9.972 -2,6560 1.28458 -36.382 -8.6aa OTII CROP PRODUCED -0.971 -1.014 1.499 -1.355 0.528 -2,369 0,5730 -0.09949 1.101 -2,468 GNP OEFI..AfOR -3.705 0.242 27,986 1.974 24 281 2.216 6.9864 0 41562 31.268 2.632 PRICES Of RICE -1.666 2 168 58.672 0,756 57 206 2 914 12.7988 -0.38196 70.005 2.532 Wiltz A f -2.112 2.916 67.959 4 034 65.847 6.960 15 2044 0.23772 81.051 7,tas COA11SE cEarAl S -17 067 -6,127 -28.689 6.464 -45.756 0.327 -2.5332 3.00600 -48.289 3.333 OTHLR CROPS -6.005 -0.526 28 669 2 .4 15 22.584 I,fla9 7.a993 0.73028 30,483 2.619 REAL. WAGE RATE 12.866 13.420 15.678 12.558 2B.534 25.978 0 8690 0.09530 29.'403 26.074 LABOR EMPLOYMENT -4.826 -4 . 699 -3.305 -4. 13a -B 13 1 -8.837 0.2293 0.02466 -7.902 -8.812 w REAL WAGE BILL 8.030 B.722 12,373 8.420 0.403 17.142 1.0983 0.11994 21.501 17 262 REAL RESIDUAL PROFITS -28.075 -19.995 42.608 -14.036 14,613 -34.031 15.2459 0.95056 29.859 -a3,080 REAL PER CAP. INCOME OF ; RURAL LANDLESS HOUSE11OLDS 14.911 13 508 3.957 12.025 1 fl. B68 26.533 -2,2001 -0-13837 16.668 2B. 394 SMALL FARM HOUSEHOLDS 11.694 11,773 13.569 11.272 25,163 23.046 0 6048 0.04607 25.768 23.092 MEDIUM FARM 11OUSEHOLDS 6.235 7 .252 15.216 7.82 1 21 .452 15,073 2.010S 0.14901 23.462 16.222 LARGE FARM imusEimi-Ds - 1.890 1.123 23.928 3,387 22,039 4 510 5.6565 0.37 154 27.595 4.882 URBAN IAOUSI711OLDS 7.633 4,800 15.545 3.874 -8 012 8.674 4.3012 9.22334 -3.711 17.89B 1101 IIOUSEIIOLDS 0.016 -1.784 -1.768 -0.42606 2. 190 PER CAP. CEREAL CONSUMPTION OF : RU14AL LANDLESS HOUSEHOLDS 9.949 B.431 -6.461 6.839 9-488 15 269 -1 .9770 -0.22363 7.511 15.046 SMALL FARM HOUSEHOLDS 7.578 7.006 4.249 6.095 11.827 13,101 -0.4314 -0.09535 11.396 13.006 MEDIUM FARM 11OUSE11OLDS 4,09.3 3.910 3.108 3,607 7 .201 7.617 0.0069 -0.04230 7 , 20a 7.475 LARGE 17ARM ltOUSF11OLDS -0 883 0.096 4.968 0.713 4 086 0.908 1 .3880 0.07567 5,474 O.BB4 URBAN HOUSE11OLDS 5.961 3.323 -15.785 1 .748 -9 824 5.071 2.3640 6.74668 -7.460 1 1.817 AGG. PER CAP. CEREAL CONSUMP I I ON 6.064 5. 192 -3.124 1.892 2 940 7,OH4 0.0637 1.19932 3.004 8.283 TABLE 3 TECHNICAL CHANGE SCENARIOS SIMULATIONS INCREASE IN INCREASE IN INCREASE IN INCREASE IN INCREASE IN RICE YIELDS WHEAT YIELDS COARSE CEREAL OTHER CROPS YIELDS OF ALL (40 ) (120%)_ - - YIELDS ( 120%)(420% CROPS (+I0%) 4.1 4.2 4.3 4.4 4.5 A B A B A B A B A 8 REAL NAT.PEA_CAP INC.(N.W.) 3.105 0.9722 3.460 3.632 0.579 0.720 8.790 7.141 7.967 6.232 TOTAL OUTPUT(N W.) -0.432 1.3561 6.606 7.422 1.400 1.376 11.643 12.032 9.608 11.093 QUANTITIES OF1.31 396 RICE PRODUCED 28.012 26.1912 3.267 -2.188 -0.810 -0.912 0.154 4.836 16.311 13.963 WHEAT PRODUCED -6.874 -1.6116 17.526 21.615 1.827 1.427 1.581 -0.967 7.030 10.232 C CEREAL PRODUCED 21.790 3.1368 6.011 6.705 3.840 14.345 -13.343 -17.782 9.149 3.202 OTH CROP PRODUCED -1.668 0.3869 0,075 -0501 1.118 0.265 21.838 24.127 10.681 12.139 GNP DEFLATOR -25.226 . -0.7974 -10.987 -3.415 2.979 1.306 -8.059 -0.045 -20.646 -1.476 PRICES OF RICE -77.962 -6.0062 -24.730 -7.659 24.447 12.7a9 4.597 6.956 -36.819 3.040 WHEAT -69.245 73.7192 -45.488 -21.023 22.279 11.832 9.458 11.052 -41.498 -0.929 C CEREAL 61.04 10.2995 16.749 10.237 -63.013 -38.187 -35.566 -21.330 -15.383 -19.490 Oflt CROP -IB.5Q9 0.1214 -4.123 0.872 -1.862 -1.153 -20.990 -4.177 -22.742 -2.168 REAL WAGE RATE -3.101 0.3801 -0.718 1.049 0.591 0.520 2.222 1.922 -0.503 1.936 LABOR EMPLOYMENT -0.662 0.1535 -0.010 0.504 0.330 0.290 0.898 0.767 0.278 . 0.858 1 REAL WAGE BILL -3.763 0.5336 -0.728 1.653 0.921 0.810 3.120 2.689 -0.224 2.793 REAL RESIDUAL PROFITS -47.235 2.5660 -8.828 9.695 11.113 7.114 23.827 39.047 -10.662 29.211 REAL PER CAP. INCOME OF RURAL LANDLESS HOUSEHOLDS 8.763 0.6059 4.483 2.500 0.250 0.591 5.321 2.269 9.409 2.983 SMALL FARM iOUSEFOLDS -0.624 0.7459 1.375 2.717 1.640 1.357 6.297 6.109 4.344 5.464 MEDIUM FARM HOUSEHOLDS -4.933 1.0624 0.960 3.917 2.169 1.653 9.131 10.504 3.659 8.568 LARGE FARM NOUSEHOLDS -16.671 1.4819 -0.986 6.002 4.143 2.817 13.236 18.465 -0.139 14.378 URBAN HOUSEHOLDS 17.809 0.9390 7.811 3.093 -2.841 -1.352 9.227 1.753 16.004 2.217 ROI HOUSEHOLDS 0.7182 1.935 -0.051 0.889 1.138 PER CAP. CEREAL CONSUMPTION OF : RURAL LANDLESS HOUSEHOLDS 11.380 0,2684 9.487 5.013 1.692 1.786 -1.944 0.097 10.307 3.582 SMALL FARM HOUSEHOLDS 7.081 0.4704 7.999 5.251 2.031 1.986 -1.782 2.197 7.665 4.952 MEDIUM FARM 11OUSEHOLDS 6.221 0.7042 9.126 6.568 1.665 1.783 -0.955 4.097 8.029 6.576 LARGE FARM HOUSEHOLDS 0.868 0.9350 11.667 9.799 1.097 1.385 -0.669 7.045 6.481 9.582 URBAN HOUSEHOLDS 16.644 0.2563 15.480 7.86' -1.059 -0.042 1.093 -0.820 16.079 3.627 AGG. PER CAP. CEREAL 797 2.337 9,920 5.516 CONSUMPTION 8.835 0.5178 10,705 6.785 1.097 1.393 -0.77 237 9.2,.1 TABLE 4 INVESTMENTS SCENARIOS SIMULATIONS ACCELERATED ACCELERAtED ACCELERATED IRRIGATION CAPITAL AND IRRIGATIONI&01 (+1%) MARKETING AND CAPITAL AND INVESTMENTS MARKETING (#10%) INVESTMENTS 3.3 3.2 3.3 A B A B A 8 REAL NAT.PER_CAP INO.(N.W.) 6.720 4.725 4.191 3.8173 -.28L 6.634 TOTAL OUTPUT(Nt.W. 7.130 8.089 3.702 4.0758 8.981 10.127 QUANTITIES OF : RICE PRODUCED 9.768 8.368 7.147 6.6259 13.342 11.681 WHEAT PRODUCED 7.050 9.346 4.044 8.0498 9.072 11.871 C CEREAL PRODUCED 6.859 2.506 3.384 4.4651 7.55t 4.739 OTH CROP PRODUCED 7.028 7.830 3.145 3.1414 8.s98 9.401 GNP DEFLATOR -13.430 -1.382 -4.732 -0.3143 -15.796 -I.539 PRICES OF RICE -23.988 0.746 -8.646 3.587 -28.281 1.505 WHEAT -29.696 -3.494 -9.570 -0.2631 -34.481 -3.626 C CEREAL -7.435 -10.166 -1.136 -8.2965 -13.003 -4.314 OTH CROP -14.000 -1.234 -4.404 -0.2734 -16.202 -1.371 REAL WAGE RATE 3.815 5.420 -0.682 0.0549 3.474 5.447 LABOR EMPLOYMENT 1.813 2.206 -0.099 0.0633 t.764 2.238 L .REAL WAGE BILL 5.628 7.626 -0.780 0.11382 6.238 7.685 REAL RESIDUAL PROFITS -19.896 5.479 8.769 17.4825 -16.516 14.220 REAL PER CAP. INCOME OF : 4 RURAL LANLESS HOUSEHOLDS 9.050 0.082 2.537 0.9601 10.318 5.562 SMALL FARM HOUSEHIOLD 5.531 6.348 3.0j4 3.1974 7.048 7.947 MEDIUM FARM HOUSEHOLDS 2.296 5.512 4.980 6.0236 4.786 8.524 LARGE FARM HOUSEHOLDS -4.248 6.036 7.043 10.2931 -0.726 10.18 2 URBAN HOUSEHOLDS II.881 3.261 3.387 0.5357 13.574 3.529 ROI HOUSEHOLDS 3.332 0.4709 1.568 PER CAP. CEREAL CONSUMPTION OF RURAL LANDLESS HOUSEHOLDS 8.991 4.674 3.977 1.5320 10.979 5.440 SMALL FARM HOUSEHOLDS 7.061 5.304 4.408 2.8424 9.268 6.725 MEDIUM FARM HOUSEHOLDS 5.725 4.727 5.627 4.3687 8.539 6.9t3 LARGE FARM HOUSEHOLDS 2.915 4.649 6.683 6.5951 6.206 7.948 URBAN HOUSEHOLDS 12.487 4.374 4.499 1.4278 14.737 5.088 AGG. PER CAP. CEREAL CONSUMPTION 7.728 4.819 4.93 '3.1904 10.198 6.434 Table 5 Reguireuents to Offset an Increase in Population Growth of 10% in t1l Northern WhaE egian Required Capital & Required Required Gain equired Gain in Required Cain in Required Required Lrrigatim Marketing InvestuAnt Gain in Rice Yields in M-,eat Yields Coarse Cereal Yields Oher Crop Yields Gain in All Yields A B A B A B A B -A B A B A B .To Hintain Real Per Capita Inmm 13.28 15.35 18.72 19.00 48.93 149.21 43.91 39.94 262.38 201.47 17.28 20.31 9.53 11.64 To antain Real Per Capita Incoxie of the Larfless 16.48 26.58 58.77 140.69 34.03 445.88 66.52 108.06 1192.88 457.12 56.05 119.07 15.85 45.28 To Maintain Per Capita Cereal Conamptim 7.85 10.77 12.28 16.27 13.73 200.54 11.33 15.30 110.56 74.54 N/P 44.43 6.11 9.41 To Hintain Per Capita Cereal Can-iWtm of the tailes 11.07 18.04 25.02 55.03 17.49 628.24 20.97 33.64 117.60 94.41 N/P 1738.35 9.65 23.54 To Hintain Weal Uae 5Ute 26.49 17.60 N/P 1135.36 NIP 503.00 N/P 172.83 323.80 331.36 95.58 99.81 N/P 48.05 oate NIP means not posaible. Ln TAI F 6 RAINF A1.1 FOODAID ANI) CIANGES IN RO CONDA rIONS SIMULATIONS DECI INE WHEAT AID INCREASE IN ROI OUfPUT OF IN RAINFAI PL 480 RICE WIIEAT COARSE OTHER ALL (-20%) ( #10%) 10%) I101.> CEREALS CROPS CROPS ( 1%) (410%) (*5%) 5 1 5.2 9.1 9.2 9.3 9.4 9.5 A A A 8 AlEAL, NAT.PER CAP INC :N.W,) -0 434 -0.2009 0.322 -0,389 0.396 -0.3031 -0.167 0.624 0.275 $OTAL OUTPUlIN.W,.) -0.239 -0.4232 -5.303 -0.282 -0.165 -0,2199 -0.139 .-0.264 -0.394 QUANTITIES OF RICE PO14DUCED -2.406 -2.2294 6.481 . 3.345 0.2$5 2.6073 0.501 -1.925 0.486 WIIEA1 PRODUCED 0.305 -0 2327 -5..635 -2. 19 -1 .003 •1.6518 0.280 1 .548 -0.414 CICERAL-S PRODUCED -5.085 2.0814 B.740 0.921 15.580 0.7182 -29.286 -22.098 -17.542 011 CROP PRODUCED -0.265 -0.6058 -0.412 0 373 -1.235 0.2908 2.604 0.993 1.276 GNP DEFLATOR 3.036 0.3555 -14.473 - 5 987 4.181 -5.5486 2.112 -1.659 -2.638 PRICES OF RICE 10.376 2.3276 -36.453 -5.387 -23.681 -4.1992 22.174 17.877 6.085 WitEAr 9.911 2.2947 -44.953 -50.694 -19.890 -8.3364 21.639 18.017 5.714 COA451 CERFALS -52.684 -5.9578 26.69:3 8.372 37.596 6.5262 -66.960 -62.023 -42 431 01IIErlRCROPS 1.981 -0.0747 -10,583 -0.063 0.261 -0 0493 -2.729 -9.798 -6.158 REAL WAGE HArE 0.027 -0 3091 -2.592 -0 691 -0 438 -0.5384 -0.367 -0.100 -0.722 LABOV FMI OYME.Nl -0.029 -0 129 -0 747 -0.237 -0.120 -0.1849 -0.001 0.078 -0.14 0% REAL WAcE am-1L -0.002 -0.4220 -3 339 -0 928 -0.559 10.7234 -0.368 -0.022 -0.836 REAL RESIDUA§. PROFITS 6 603 0 9702 -33.870 -6.322 -8.824 -4.9285 6.687 -0.806 -3.936 REA1. PER CAP. INCOME OF : RURAI. LANDLESS HIOUSEHOLDS -l.535 -0.2808 4.014 0.215 1.186 0. 1677 0.044 1.362 1.380 SMALL FAAM HOOSEILn.DS -0.008 0.2006 -2.060 -0.797 -0.375 -0.6250 0,809 0.799 0.306 MEDIUM I-ARM HOUSF1101DS 0.627 -0 0668 -4.986 5.1.255 -.018 -0.9782 0.962 0.268 -0.384 LARGE -ARM HOUSEHOLDS 2.236 0.2043 -12.505 -2.412 -3.054 -.8806 2.003 -0.493 -1.712 URBAN IlOUJSFHOLDS -2.438 -0.4926 9.920 1.539 3.158 0.8878 -2.804 0.948 1.095 R01 SOUS1L1O 0s 0.1497 1.228 3.428 0.9570 0. 147 3.276 3.933 PFR CAP. CEIEAL CONSOMP[ION O~f RURAL LANDI.ESS I1OUSFHIOLDS - 1 002 -0 0288 6.400 5.402 15565 1,093t 2.100 0.623 2.689 SMALL FARM HSOUSC.HOIDS -0 498 -0.0194 3. 324 0.893 1 102 0.6958 2.122 -0.087 1.866 MEDIUM FARM 11OUSEI1OLDS -0 376 0.0013 2.914 5.006 1.255 0.7843 1.666 -1.132 1.237 IARGE [ARM H10USEHOLDS 0. ,16 -O 0239 1.132 1.396 0.884 1.0884 0.207 -3.396 -0.608 URUAN 11OUSEIOliDS 2.064 -0,2987 12.387 3.009 2.628 2.3453 -0.965 -0.193 5.908 AGG. PLA CAP CEOSEAL CONSUMP iION -0.813 -0.0802 5.418 5.538 1.522 1.1987 1.043 -0.729 1.517 TABLE 7 TAXES AND INCOME REDISTIBUION SCENARIOS SIMULATIONS LAND TAX PROGRESSIVE EXCISE TAX LAND INCOME TAX EXCISE TAX LAND FROM (I0 ) RURAL INCOME ON URBAN TO RURAL 40 RURAL TO RURAL RICH TO I TA XOODS POOR POOR POOR POOI 6.1 6 2 6 3 7.1 7.2 7.3 0.1 A B A B A B A A 0 A N A a REAL NAT.PER_CAP INC.IN.W.) -1.262 -1.593 -4,750 5.972 -1,4539 1 5582 0 127 0 637 -0 690 -0.368 -2 350 -1.017 1.663 2 3929 TOTAL OUTPUTIN W ) 0376 -0 0.17 -1.400 -0 138 -0 3405 -0 2069 0 537 0.030 0 313 0 007 0.294 -0920 0.775 0.0621 QUANII:IESPOF :-,0 033 RIE PROOUCFO 0 900 0.232 3 392 0 883 0 9853 0 8401 -1.090 -0 190 -0 4098 -0 020 0.405 3 317 -4.704 -0.3831 WHEA1 PHODtUCED -1 069 -0.161 -3.997 -0.574 -1 4668 -0 Gals I 668 0 155 I 007 0.060 -0 088 -4 444 2.266 0.2683 C CEREAL PRODUCED 3 168 -0.231 11.941 -0 806 2 .790 7 9104 -4 712 -0.494 -2 633 -0.263 -3.066 42.928 -6.701 0.1723 0141 CROP PRODUCED -0 435 0.027 -1 628 0.099 0 2775 -0 6664 0 529 -0 003 0 260 0 010 0 868 -3.615 0.858 -0.0325 GNP OEFLATOR -4.769 -0 321 -17.743 -1 206 -0 9393 0 0696 6 936 0 312 4.405 0.104 21.846 2.022 9.929 0.5763 PRICES Of RICE -40 343 -0.308 -36.507 -1.198 -6.6472 -7 5170 18.311 0 883 12 136 0,633 26 072 -37.893 23.409 0.8800 WIEAr -11.911 -0.776 -44.464 -2 .78 -8.4080 -9 4329 11 464 1 l06 1 1201 0.531 23 369 -45.9.24 25 391 1.5692 C CEREAL 5 778 -O 383 21 916 -4.276 8 4902 20 74.10 -10 245 -0.751 -6.278 -0.785 10918 111.399 -13.088 0.0724 I Olit CROP -4.764 -0.359 -47.687 -t.336 -4.4530 0 4463 6 784 0.166 2 992 -0.046 17.922 4.046 9.283 0.5508 U REAL WAGE RATE -0.649 -0.067 -2.304 -0.253 -0 8908 -0 6316 t 001 0 066 0 633 0.022 -4.t61 -3.020 1.339 0 ita9 LABOR EMPLOYMENI -0.169 -0.020 -0 630 -0 077 -0,2530 -0 2105 0 253 0.019 0.450 0.006 -0.391 -1 018 0.3530 0.0355 REAL WAGE BILL -0.788 -0 088 -2.934 -0 330 -1.1439 -0 8422 1.253 0.085 0 783 0.028 -1.552 -4.037 4.6923 0 1544 REAL RESIDUAL PROFITS -20.656 -10.807 -39.328 -3.049 -9.096 -7.5327 -27.009 -44.002 8.409 0.215 10.352 -36.022 24.6072 1.4240 REAL PER CAP INCOME OF RURAL IANOLESS HOUSEHOLDS 4 280 -0.090 5.395 0.314 *O.0285 -0 5253 27.062 29 193 28.634 29.969 21 667 26.771 12.2106 15.1457 SMALL FARM 11OJUSEHOSLD -, 635 -1.226 -1.759 0 227 -I 2104 -1,3037 -4 325 -4.8t9 0.248 0.010 -3.948 -6.647 0.8906 0.4013 MEDIUM FARM IIOUSEIiOLOS -3 893 -2 610 - 5.280 -10,499 2 2161 -2 030 -8 458 -10.233 -II 788 -42.793 -4.006 -10.328 2.8419 0.2272 LARG FARM 11OUSEliOLOS -8.544 -5 008 -34.306 -21 449 -4 1995 -3 5438 -44 256 -19 364 -22 657 -26 660 -0 961 -17.336 -0 0844 -7.4867 Uf1BAN 11OUSE1lOIS 3.303 0.004 13.091 0 840 0 4927 -0 8543 -5 762 -1.095 -2 771 -0.061 -24 699 -5.653 -7 1882 -0.3844 904 1IOUSE1OLDS 0,272 4.8 I 1004 -0.267 -0,996 -6.736 -0.4920 PER CAP. CEREAl. CONS OF : RURAL LANDLESS H1OUSE1IOtDS 1 327 0.026 5 320 0 607 0 5018 -0 4062 I5 624 18.309 16.903 48 850 13 484 16.034 6.2437 9.3951 SMAlL FARM 1USEUOLDS -0 290 -0.658 I 528 0 487 0 0294 -0 6832 -4 445 -3.061 - .273 -0.075 -3 077 3.966 I 3583 -0.1176 MEDIUM FARM ItOUSE1OLDS -l 240 -I.3A4 -6 079 -6 59b . -0 2621 -0 7792 -7 094 -6, t7 -8 425 -7.440 -3 34t -4 391 -0 8761 -0,0927 LARGE FARM ifoUSEitOjIOS -3,240 -2.688 -13.424 - 44000 -0 7876 -0 6815 -40 578 -40.871 -14.266 -44 t63 -1 248 -3,695 -3.4037 -4.4015 URBAN IiOUSEHOILDS 3.208 0.463 12.563 1 185 0 4008 -0 3200 -5 566 -0.968 -2.868 -0.087 -16.824 -3 312 -6.9867 -0.6610 AGG. PER CAP. CEREAL C 0usUM P 11N 0 1440 -0 806 0 436 2 958 0.0203 0,6794 -0 042 14.986 0.087 1.560 -0.363 2.263 -0.41474 2. 4280 - 38 - Taxation and Income Distribution Scenarios Sinulations 6.1 to 6.3 in Table 7 introduce various forms of taxation purely as revenue measures. Income is taken from the rural S-ctor and used for unspecified purposes which do not affect agricultural demand or supply.- In scenario 6.1, a land tax is levied at a rate of 10% of the land rental income (not of rents received by landlords). In scenario 6.2, a progressive rural income tax is levied, a scheme which does not exist at present. A rate of 10% and 20% is levied on nominal incomes of medium and large farmers respectively. Small farmers and the landless are untaxed. Scenario 6.3 imposes an excise tax on nonagricultural goods. In the model, this is achieved by an exogenous increase of the price index of nonagricultural goods of 5%. This would obviously translate into higher rates of excise taxaticn for taxed commodities, as th-e list of goods and services would not cover all nonagricultural commodities. Unlike the land and income tax, the excise tax falls also on the urban group. Simulations 7.1 to 7.3 are income redistribution schemes in which the nominal per capita incomes of the rural poor are given a boost of 30% which is financed out of different sources of tax revenues. In scenario 7.], the revenue source~is a land tax just sufficient to finance the welfare payemnts, i.e., no leakages are assumed. Even then, the required land tax rate is very large, nearly 42% of the land rental rate. In scenario 7.2, such a progressive income tax is used to finance a welfare payment that increases the nominal incomes of the landless by 30%. The required income tax rates are 12.82% and 25.64% for medium and large farmer groups respectively. - 39 - In scenario 7.3, an excise tax at the rate of 26.65% is used to finance the welfare payment of 30% to the landless group. In version A of simulation 6.1, a land tax alone translates into reduced demand for agricultural output, substantial price drops (GNP deflator is -4.8%) and a decline in real residual profits of 20.6%. Urban consumers and the landless gain while producers lose. In version B all price effects are dampened and therefore land rents decline just a little more than the initial tax bite of 10%. Qualitatively, the effects of the progressive rural income tax are very similar to those of the alnd tax, but quantitatively the effects are larger as the tax bite is bigger in absolute magnitude. The excise taxes have a more even incidence, as it falls also directly on the landless and urban groups. Nevertheless, large farmer groups are most heavily penalized, as real land rents decline by somewhat more than 7% in both versions A and B. Quantitatively, scenarios 7.1 to 7.3 are comparable as in each case a sufficient tax is levied to finance equal initial increases in incomes of the rural landless. In all three scenarios, the income distribution con- sequences are better when price effects are small (version B) than when they are large (version.A). When land and income taxes are levied, the landless receive at least 27% of the initial income boost of 30%. However, under the excise tax scenario, they receive only a 21.6% net gain. The land and income taxes are largely financed by farmer groups whereas the urban group and the ROI finance a substantial proportion uTider the excise tax scenario. By com- paring the-;:eal per capita income effects of NW as a whole across scenarios, we also see that the land tax is the most efficient income transfer mechanism, - 40 - as real per capita income even increases albeit slightly in both versions A and B. Income taxes result in modest aggregate income losses while under the excise tax, the aggregate income loss is 2.35% and 1.02% in versions A and B respectively. Note also that for the land or the income tax scenarios, the income losses of large farmers are about the same under versions A and B. But in the excise tax scenario, large farmers lose substantially only under version B. The income loss imposed on ROI consumers sharply reduces wheat and rice prices and increases coarse cereals prices, leading to a reduction in land rents in NW which produces little coarse cereals. Scenario 8.1 achieves income distribution objectives by a direct transfer of land from large farmers to the landless. Aggregate NW per capita income increases in both versions A and B. The landless achieve an income gain of 12.2 and 15.1% respectively. In version A, large farmers lose only 0.08% of their income. The reason is that increased food demand of the land- less drives food prices up sufficiently to lead to an increase in land rents which nearly compensates for the land given away. The real losers are urban consumers. In version B, food prices barely rise and large farmers now bear the real burden while urban consumers suffer little. By comparing scenario 7.1 to 8.1, we can see that it is by no means easy to raise landless income by say 25%. The land tax or the land reform would have to transfer about 40% of land rent or of land itself to achieve these results. -41- CONCLDING REMARKS The results of the scenarios do not lend themselves to straight- forward summarization; the complexities of the interactions modelled are reflected in a wide variety of outcomes which depends on the type of inter- vention, investment or technical change, the commodity directly affected and on the trade regime pursued. The theoretical modelling exercises indeed implied such complexities. Nevertheless, a few major points stand out. First is th eimportance of income effects, both via producer incomes as well as via the income effects of food price changes. The income effects lead to several conclusions which a model containing price effects alone would not predict. Income effects influence the distributional outcome of technical change. Under price-inelastic demand one would expect all pro- ducer groups to lose invariably from technical change. However, when income effects are incorporated and are large (e.g., technical change in the "other crops" aggregate), large farmers experience a gain in real income even when the NW region is completely closed. In the open economy case, income effects also prevent the ROI from sharing in the gains of reduced population growth in the NW region. Gains in real incomes of poor NW groups increase agricul- tural demand sufficiently to prevent a price decline which could benefit ROI consumers. Food aid also appeafs less of a disincentive to production than normally assumed once income effects are introduced. In version B, for example, NW wheat output declines by only a fifth of what is erstwhile provided in food aid; and even this is largely compensated for by increases in the supply of all other agricultural commodities. We also note that the poorest group, landless workers, gain sub- stantially from reduced population growth and faster urbanization, or from any combination thereof - after all labor becomes scarcer. More surprising is that they do not lose under any of the technical change or investment - 42 - scenarios. Real wages appear to be relatively unaffected by the technical changes, quite unlike real land rents (residual profits). Wages are deter- mined by the interplay of demand and supply while land rents are determined residually. The rents appear to be much more volatile than wages, despite the fact that both the demand and supply of labor are rather price inelastic. We note, however, that laborers gain more from technical changes and agricul- tural investments when NW does not trade at all (version A), i.e., income effects from price changes are more important than those associated with ex- panded labor demand. For the landowners on the othrr hand, trading opportuni- ties with the res of India are vital. The trading possibility sharply reduces their losses from technical changes and investment programs or sharply increases their gains. Thus, an important conflict of interest about trade exists not merely between rural and urban groups, but also between landless workers and farmers. We stress again the preliminary nature of the findings in this paper. They demonstrate the flexibility and power of the model we have developed. But we have yet to do counterfactual analysis. And the policy analysis can be made more precise. At the present time we do not, for example, consider the cost of distributive mechanisms. But if these costs can be estimated, the model could be used to simulate the outcome of alternative allocations of a fixed budget to different programs. Furthermore, the set of investments con- sidered is rather limited to those for which we have econometrically estimated coefficients. But the list of investments can be expanded. The direct output and factor demand effects of any investment can be estimated, as is done in benefit-cott analysis for example. Its effects can be modelled straightforwardly with the tools at hand. -43- Appendix I: MIGRATION ELASTICITIES FOR NORTH INDIA Appendix Table I.1 lists six migration streams for which Sanjay Dhar's study (1980) provides elasticities with respect to the rural wage. Inside and out- side refer to within and outside of a state. Let the rural groups of the region studied be denoted as k=1, . . . K-1, let k=l denote the poorest group, and let K stand for the urban group. We will assume that all immigrants within the rural population join the poorest group 1, i.e., they do not bring any assets with them such as land. Therefore the migrant streams 4, 5 and 6 in Appendix table 1 (from urban areas within and outside the region and from outside rural areas) will be added to the poorest group. Let AN refer to the change in IMIG population in the first class via immigration and E as the corresponding immigration elasticity IMIG 3N 1 DUlRl U2R1 3R2R1 aw aw 3W aw Normalizing by N1 leads to 9NIMIG IMIG 1 U1R1 W UlRl 3U2R1 W U2R1 (2) 1 -- + - 1 aW N 9W UlR1 N1 @W U2Rl N1 @R2Rl W R2Rl - 3W R2R1 N1 On the other hand, a change in the wage rate may affect the out-migration behavior from the region of all classes equally. If marginal returns to labor within the region change, it may speed up or slow down out-migration of all residents, irrespective of their income or asset ownership class. We therefore assume that the first three migrant streams in table 1 (to urban areas inside and outside the region, and to outside rural areas) come evenly distributed from all rural classes. -44 This implies that (3) E E = EMIG k=l, . . K-1 k R where EEMIG is the overall rural emigration elasticity. Now let the total R. K-1 rural population be denoted as NR = E N.. Then k=l ENIG (4) ~NR R1U1 .=RlU2 + RlR2 aW - aw aw 3W where ANR EMI refers to the change in rural population via emigration. Converting to rates of changes by normalizing by the total rural population NR EMIG aRlUl W RlUl + LR1U2 W RlU2 + 3R1R2 W RlR2 (5) c_-_ _+_ 1 R aW R1U1 NR 3W RlU2 NR aW RlR2 NR The migration elasticities for the rural groups are (6) EM = EIMIG + ER m 1. R k = EIG k=2, . K-1 For the urban group (k=K), 3NK ;RlU1 U1R1 aw -aw - aw i.e. we add the streams 1 and 4 in Appendix Table I.1 (the within region rural to urban flow less the change in the within region urban to rural flow). Normalizing by NK leads to NK W UR1U W RlUl 9U1R1 W UlRl (8) inK =W N aw R1Ul N 3W UlR1 N K K K Appendix Table 1.2 gives the final elasticities, the Emk' for each of the groups used in this study. Weights used to compute these elasticities are from Dhar (1980)'s reported estimates of actual migrants by migrant stream and from estimates of the urban and rural population in the Census of India, 1971. Appendix Table 1.1: Migration Elasticities 0-1 Year 1-4 Year Elasticity Migrant Stream Elasticity Source Elasticity Source (1) ZRIU1 W1 Inside rural to -4.8 1/ -7.7 1/ Wl RiUl inside urban (2) RIU2 W1 Inside rural to - .846 Table 3.4 -1.35 Table 3.4 u41 RIU2 outside urban (3) 31lR2 W1 Inside rural to - .19 Table 3.7 0 Table 3.7 21 1 I R1R2 outside rural (4) ;UR1 1 Inside urban to 1.71 Table 5.2 1.34 Table 5.1 Wl UIR1 inside rural (5) U2R W1 Outside urban to .371 Table 4.6 .050 Table 4.5 Sl U2RL inside rural (6) aR2RI W1 Outside rural to .0925 Table 3.7 0 Table 3.72/ 4W1 R2R1 inside rural Not required Inside rural to .63 Table 5.4 .861 Table 5.3 inside rural 1/ Not estimated by Dhar (1980). The rural to rural migration elasticity within state (0.63) exceeds the one from outside the state (0.0925) by a factor of 6.8. Similarly, the urban to rural migration elasticity from inside the state (1.71) exceeds the one from outside the state by a factor of 4.5. The average of these factors is 5.7 and we apply this factor to the inside rural to outside urban elasticities to obtain the inside rural to inside urban elasticities shown here. 2/ Dhar's estimate is of the wrong sign, but not significant and we set it to zero. - 46 - Appendix Table 1.2 Migration Elasticities (E mk) by Producer Group Groups Marginal Small Medium Large Urban Migration elasticity Emk .115301 .100161 .100161 .100161 -.404504 - 47 - Appendix Table 1: Agroclimatological Regions and the States and Union Territories of India that Comprise Them-1 Agroclimatological Region State/Union Territory Districts Semi-Arid Tropics (SAT) Andhra Pradesh Adilabad, Nizamabad, Karimnagar Medak, Warangal, Mahb(ibnagar, Hyderabad, Nalgonda, Khammam, Kurnool, Guntur, Vishakapatnam Anantapur, Cuddapah, Ongole, Nellore, Chitoor. Gujarat All Karnataka Bidar, Gulbarga, Bijapur, Belgaum, Dharwar, Raichur, Shimoga, Bellary, Chikmagalur, Chitradurga, Hassan, Tumkur, Mandya, Mysore, Bangalore, Kolar. Madhya Pradesh All .Maharashtra All Rajasthan All Tamil Nadu Dharmapuri, The Nilgiris, Coimbator Salem, Tiruchirapalli, Pudukkottai, Madurai, Ramanathapuram, Tirunelvel Dadra & Nagar Haveli Eastern Rice (ER) Arunachal Pradesh All *Assam All Bihar All Manipur All Meghalaya All Mizoram All Nagaland All Orissa All Tripura All Uttar Pradesh Jalaun, Jhansi, Hamirpur, Banda Fatehpur. Rae Bareli, Sultanpur, Faizabad, Basti, Allahabad, PTatap- garh, Jaunpur, Azamgarh, Gorakhpur, Mirzapur, Varangsi, Chazipur, Ballia, Deoria. West Bengal All - 48 - Agroclimatological Region State/Union Territory Districts Coastal Rice (CR) Andhra Pradesh Srikakulam, East Godavari, West Godavari, Krishna. Goa Karnataka North Kanara, South Kanara, Coorg. Kerala All Pondicherry ----- Tamil Nadu Chingliput, North Arcot, South Arcot, Thanjavur, Kanyakumari. Northern Wheat (NW) Chandigarh Delhi Haryana All Himachal Pradesh All Jammu & Kashmir All Punjab All Uttar Pradesh Dehradun, SaharanDur, Bijnor, Nainital, Muzaffarnagar, Meerut, Moradabad, Rampur, Bulandshahr, Budaun, Bareilly, Pilibhit, Mathura, Aligarh, Agra, Etah, Mainpuri, Farukhabad, Shahjahanpu Kheri, Etawah, Hardoi, Sitapur, Kanpur, Unnao, Lucknow, Barabanki Bahraich, Gonda. For those states that fall into two agroclimatological regions, districts are allocated and'identified individually. Appendix Table 2: PRODUCER CORE PRICE ELASTICITIES FOR NORTH INDIA BEFORE AND AFTER CONVEXITY ADJUSTMENT Rice Wheat Coarse Cereals Other Crops Fertilizer Labor Before Convexity Price Elasticities Rice 0.2799 -0.5095 -0.0765 0.3878 0.0342 -0.1040 Wheat -0.0863 0.3443 0.0833 -0.1380 0.0082 -0.2405 Coarse Cereals -0.0807 0.5189 0.4826 -0.6051 -0.0039 -0.4443 Other Crops 0.0418 -0.0878 -0.0618 0.2335 -0.0451 -0.0712 Fertilizer -0.0321 -0.0457 0.0035 0.3930 -0.1226 -0.1951 Labor 0.0074 0.1013 0.0301 0.0471 -0.0148 -0.1886 After Convexity Pripe Elasticities Rice 0.3545 -0.5095 -0.0765 0.4252 -0.0150 -0.1787 Wheat -0.0863 0.3703 0.0862 -0.1380 0.0082 -0.2405 Coarse Cereals -0.0807 0.5367 0.5972 -0.6051 -0.0039 -0.4443 Other Crops 0.0458 -0.0878 -0.0618 0.2335 -0.0451 -0.0846 Fertilizer 0.0141 -0.0457 0.0035 0.3930 -0.1698 -0.1951 Labor 0.0127 0.1013 0.0301 0.0560 -0.0148 -0.1852 I/ Elasticities are computed at base year 1973-74 prices and quantities. - 50 - Appendix Table 3: Price and Income Elasticities of Demand, by Commodity and by Producer Group, North India Commodity Elasticities With Respect to the Prices of Producer Groups Commodities Rice Wheat Coarse Cereals Other Food Nonfood Income Marginal Rice -0.8506 0.2967 -0.1529 0.5095 0.2019 0.7154 Wheat 0.7634 -0.1234 0.0403 0.1282 -0.2040 1.0209 Coarse Cereals -0.4305 0.0390 -0.5068 0.6870 0.2159 -0.4536 Other Food 0.3827 0.0373 0.1902 -0.7975 0.1918 1.1351 Nonfood 0.2463 -0.0978 0.0988 0,3105 -0.5532 1.5761 ill1 Rice . -0.8708 0.3068 -0.1784 0.5257 0.2143 0.6864 Wheat 0.7591 -0.7219 0.0287 0.1304 -0.1987 1.0034 Coarse Cereals -0.5425 0.0285 -0.4605 0.7426 0.2295 -0.5725 Other Food 0.3667 0.0366 0.1781 -0.7889 0.2051 1.1181 Nonfood 0.2331 -0.0889 0.0881 0.3182 -0,5530 1.5760 Kedium Rice -0.9220 0.3362 -0.2436 0.5690 0.248 0.6074 Wheat 0.7438 -0.7190 0.0030 0.1381 -0.1815 0.9619 Coarse Cereals -0.8321 0.0026 -0.3363 0.8869 0.2634 -0.8598 Other Food 0.3281 0.0358 0.1504 -0.7678 0.2378 1.0781 Nonfood 0.2011 , -0.0675 0.0641 0.3374 -0.5506 1.5740 Large Rice -1.0286 0.4318 -0.3824 0.6778 0.2991 0.3922 Wheat 0.6721 -0.7216 -0.0169 0.1744 -0.1103 0.9009 Coarse Cereals -1.4866 -0.0469 . -0.0127 1.2177 0.3264 -1.3146 Other Food 0.2631 0.0430 0.1221 -0.7282 0.2976 1.0253 Nonfood 0.1466 -0.0353 0.0421 0.3717 -0.5274 1.5520 Urban Rice -0.9499 0.3612 -0.2799 0.5975 0.2590 0.5512 Wheat 0.7251 -0.7197 -0.0216 0.1476 -0.1629 0.9459 Coarse Cereals -1.0034 -0.0103 -0,2516 0.9734 0.2799 -0.9788 Other Food 0.3111 0.0377 0.1430 -0.7574 0.2535 1.0643 Nonfood 0.1868 -0.0591 0.0584 0.3464 -0.3445 1.5682 -51- Appendix Table 4 SHARES IN TOTAL CONSUMPTION BY COMMODITY AND BY PRODUCER GROUP, NORTH INDIA Commodities Producer Group Rice Wheat Inferior Other Non-food Cereals Crops Marginal 0.3494 0.2580 0.4475 0.2257 0.1897 Small 0.0602 0.0442 0.0708 0.0414 0.0368 Medium 0.3021 0.2651 0.3108 0.2555 0.2427 Large 0.1168 0.1979 0.1218 0.1708 0.1793 Urban 0.1715 0.2348 0.0491 0.3067 0.3514 Total 1.0000 1.0000 1.0000 1.0000 1.0000 - 52 - Appendix Table 5 SHARES OF COMMODITIES IN CONSUMPTION BY PRODUCER GROUP, NORTH INDIA Commodities Producer Group Rice Wheat Inferior Other Non-food Total Cereals Crops Marginal 0.1591 0.1681 0.0727 0.3828 0.2173 1.0000 Small 0.1522 0.1599 0.0639 0.3896 0.2344 1.0000 Medium 0.1283 0.1611 0.0471 0.4024 0.2593 1.0000 Large 0.0763 0.1850 0.0284 0.4156 0.2947 1.0000 Urban 0.0672 0.1317 0.0069 0.4479 0.3464 1.0000 All Groups 0.1108 0.1585 0.0395 0.4126 0.2786 1.0000 - 53 - Appendix Table 6 SHARES IN THE TOTAL SUPPLY 6F AGRICULTURAL INPUTS BY PRODUCER GROUP, NORTH INDIA Agricultural Inputs Producer Group Agricultural Bullocks Agricultural Agricultural Labor Land Implements and Machinery Marginal 0.4256 0.0459 0.0192 0.0570 Small 0.0699 0.0629 0.0216 0.0322 Medium 0.3341 0.5082 0.3450 0.4067 Large 0.1465 0.3588 0.5635 0.4509 Urban 0.0239 0.0241 0.0507 0.0532 Total 1.0000 1.0000 1.0000 1.0000 -54- Appendix Table 7 SHARES IN TOTAL INCOME OF INCOMES FROM AGRICULTURAL INPUTS, BY PRODUCER GROUP, NORTH INDIA Agricultural Inputs Producer Group Agricultural Bullocks Agricultural Agricultural Total Labor Land Implements Agricultural and Machinery Income Marginal 0.5969 0.0175 0.0174 0.0282 0.6600 Small 0.5621 0.1433 0.1167 0.0951 0.9171 Medium 0.3499 0.1543 0.2479 0.1599 0.9120 Large 0.1768 0.1266 0.4704 0.2060 0.9798 Urban 0.0146 0.0043 0.0214 0.0123 0.0525' All Groups - 0.2631 0.0751 0.1779 0.0973 0.6133 a/ Includes income from rural non-agricultural wages. - 55 - Appendix Table 8 SHARES IN TOTAL REAL INCOME BY PRODUCER GROUP, NORTH INDIA Marginal Small Medium Large Urban Total 0.2420 0.0441 0.2638 0.1654 0.2847 1.0000 SHARES OF AGRICULTURAL COMMODITIES IN VALUE OF TOTAL AGRICULTURAL OUTPUT, NORTH INDIA Rice Wheat Inferior Cereals Other Crovs Total 0.0584 0.3447 0.0554 0.5416 1.0000 SHARES OF AGRICULTURAL INPUTS IN TOTAL COST OF PRODUCTION, NORTH INDIA Agricultural Bullocks Agricultural Agricultural Total Labor Land Implements and Machinery 0.4233 0.1237 0.2928 0.1602 1.0000 SHARES IN TOTAL POPULATION BY PRODUCER GROUP, NORTH INDIA Marginal Small Medium Large Urban Total 0.3273 0.0538 0.2569 0.1127 0.2493 1.0000 Appendix Table 9: Output Supply Elasticities With Respect to Exogenous Shifter Variables RAIN HYV IRK ROADS LAND 2/ CAPITAL Rica 0.1163 0.0818 0.6881 -0.0568 0.0893 0.5944 Wheat 0.0177 0.4070 1.2331 -0.0044 0.0893 0.5944 Cereals -0.2282 -0.0783 1.1583 0.3456 0.0893 0.5944 Crops 0.0466 -0.0348 0.7762 -0.3403 0.0893 0.5944 Fertilizer 0.2503 0.9824 2.0235 -0.2538 3/ Ln Labor - 0.0085 0.0558 0.3166 -0.0567 0.4130 0.0902 Bullocks 0.0085 0.0558 0.3166 -0.0567 0.5824 0.0437 1/ Unless otherwise indicated, all estimates are from Evenson (1981) and uses 1973-74 base year quantities of the variables. 2/ These estimates are from Evenson and Binswanger (1981). Since only the elasticities of aggregate output output with respect to both land and capital are available from this study, these estimates are assumed to be the same across all crops. 3/ Except for the last two columns of these rows, these estimates are computed to be a third of the value weighted sum of the elasticities in the respective columns. -57- List of Symbols Used B = bullock services E!= I = technology shifters; i.e.; shifts in output supplies S OT Q i @t and factor demands for given fixed input levels. These are profit function definitions. G square matrix of elasticities and shares. K*= column vector of exogenous shifter variables L = labor services M = total iominal income m = real per capita income MN = nonagricultural income N = population NA = nonagricultural commodities NR = rural population NU = N = urban population K P = output prices P = output price indices Q = [Y,-X] = vector of outputs and (negative) variable inputs S = rent to fixed factor, usually land s. = share of output i in total revenue or share of factor i in total cost 1 t = time U' = column vector of endogenous variables V = [P,W] = vector of output and variable input prices W = wage rate or variable input prices w = real wage rate X = variable inputs Y = outputs - 58 - y = per capita output Z = fixed factors;.Z1 refers to land H* = variable profits T = technology index Modifiers of variables unless already defined above: X = level X,XT = a column and a row vector of the X variables respectively dX 1_ X' = dt X= total rate of change (n growth rate) of variable X with respect to time X* = exogenous component of the rate of change of a variable (except that H* stands for maximized variable profits). Indices and sets of inputs and outputs: g = shifter variables i = commodities (outputs, inputs) j = commodities (outputs, inputs) k = income groups K = set of income groups or their total number 0 = set of outputs or total number of outputs I = set of inputs or total number of inputs VI = set of variable inputs or their total number Definitions of parameters: a: commodity demand elasticities 6: output supply and factor demand elasticities from the profit function E: factor supply elasticities - 59 - Shares and proportions: 6 ik share of factor i in the income of income group k ik X. = proportion of factor i supplied by income group k, but also refers to -ik the 'proportion of commodity i consumed by income group k. Nk = share of income group k in the population pik = share of commodity i in the total expenditures of income group k Vk = proportion of real income accruing to income group k. 60 - References Binswanger, Hans (1983). "Variable Profit Functions and Definitions of Rates and Biases of Technical Change", World Bank, mimeo. Binswanger, Hans, Jaime Quizon and Gurushri Swamy (1982). "The Demand for Food and Foodgrain Quality in India", World Bank - ARU Discussion Paper No. 6, November 1982. Dhar, Sanjay (1980). An Analysis of Internal Migration in India. Unpublished Ph.D. dissertation. Yale University. Evenson, Robert (1981). "Green Revolution in North Indian Agriculture: An Ex-Post Assessment of Economic Effects", Economic Growth Center, Yale University, mimeo. Evenson, Robert and Hans Binswanger (1981). "Estimating Labor Demand Function for Indian Agriculture", in Contractual Arrangements, Employment and Wages in Rural Labor Markets in Asia, Hans Binswanger and Mark Rosenzweig (eds.), forthcoming. Handal, D.S. and D.K. Grover (1976). "Input Prices, Production and Profitability in Haryana", Indian Journal of Agricultural Economics, July - September 1976. Quizon, Jaime and Hans Binswanger (1983a). "Income Distribution in Agriculture: A Unified Approach", American Journal of Agricultural Economics, forthcoming (August 1983). Quizon, Jaime and Hans Binswanger (1983b). "Factor Gains and Losses in the Indian Semi-Arid Tropics: A Didactic Approach to Modeling the Agricultural Sector", World Bank - ARU Discussion Paper No. 10. Revised May 1984. - 61 - Rosenzweig, Mark (1980). "Neoclassical Theory and the Optimizing Peasant: An Econometric Analysis of Market Family ,Labor Supply in a Developing Country", The Quarterly Journal of Economics, February 1980. Singh, Inderjit (1981). Small Farmers and the Landless in South Asia. World Bank, Development Economics Department, mimeo. 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 aud 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. Reoort 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. Resort 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, September 1983, Revised May 1984. Reuort No.: ARU 10 The Distribution of Income in India's Northern Wheat Region by Jaime B. Quizon, Hans P. Binswanger and Devendra Gupta, Augusc 1933. Revised June 1984. Report No.: ARU 11 Population Density, Farming Intensity, Patterns of Labor-Use and Mechanization by Prabhu L. Pingali and Hans P. Binswanger, September 1983. Reoort 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 Equi.table 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.: ARUJ 15 Patterns of Agricultural Protection by Hans P. Binswanger and Pasquale L. Scandizzo, November 15, 1983. Report No. : A.RU 16 Factor Costs, Income and Supply Shares in Indian Agriculture by Ranijan Pal and Jaime Quizon, December 1983. Report No.: ARU 17 3ehaviorai and M,aterial Determinants of Production Relations in Land Abundant Tropical Agriculture by Hans P. Binswanger and John HcIntire, 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. ReEort 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. PingalIi, October 1984.