Turkey agricultural sector model : further results for the livestock sub-sector

This internal working paper is prepared for STA1FF USE ONLY. The view expressed are not necessarily those of the World Bank. TURKEY AGRICULTURAL SECTOR MODEL FURTHER RESULTS FOR THE LIVESTOCK SUB-SECTOR AGREP Division Working Paper No. 68 Prepared by: M.C. Evans and V. Le-Si (consultants) Economics and Policy Division Agriculture and Rural Development Department March 1983 ACKNOWLEDGEMENTS We would like to thank R. Burcroff for his help, suggestions and comments. We would also like to acknowledge the secretaries of AGREP for their patience and typing skills in putting this paper together. I. INTRODUCTION 1.01 This report is about further development and application of the Turkey Agricultural Sector Model (TASM). The use of TASM by the' Turkey Agicultural Sector Study Mission in May-July 1982 is described in Report No. 4204-TU (Turkey - Agricultural Development Alternatives for Growth with Exports, January 1983). 1/ This analysis indicated that Turkey has an actual or potential comparative advantage as an exporter in the case of the majority of agricultural commodities currently being produced. Continuing liberalization of trade policy towards an essentially "free trade" position with respect to agriculture combined with attainable incremental investments in extending the irrigated crop area, increasing livestock inventories and improving both crop and livestock productivity, would ensure a minimum feasible rate of growth for agriculture of about 2% p.a.2/ between 1979 (the base year for TASM) and 1990. The total area of cultivated land would decrease by 10% and the intensity of working capital use in agriculture (fertilizer and tractor power) would increase. Agricultural employment would not expand fast enough to keep up with the growth of the agricultural labor force (projected at 1% p.a.). The composition of agricultural output, measured at 1979 prices, would not change greatly in terms of the major commodity groups, but within the grain groups there would be a substantial increase in the production of coarse grains (especially barley) relative to the production of wheat. 1.02 The value, at 1979 prices, of agricultural exports projected for 1990 by TASM was more than four times the level for 1979, and the commodity / Technical aspects of TASM are discusssed in V. Le-Si, P.L. Scandizzo and H. Kasnakoglu's "Turkey Agricultural Sector Model", The World Bank, March 1983. 2/ Projected agricultural GDP growth at 1979 prices of 1.6% p.a. adjusted for expected movements in the barter terms of trade between the agricultural and non-agricultural sectors (See Report No. 4204-TU, Annex 5, paras. 24 and 41). -2- composition of this trade was shown as being substantially different in these two years. The grain, fruit and nut and vegetable groups would have larger shares in total agricultural exports in the future and the "industrial" crop group (mainly cotton and tobacco) a much smaller share than before. However, because of the effect on foreign exchange requirements of the increased intensity of input use in agriculture, the net foreign exchange inflow into the sector was projected to be much smaller in 1990 than in the base year under an assumed free trade regime. I! A policy of foreign trade liberalization would mean that the inter-sectoral barter terms of trade would be more in agriculture's favor in the future than in the recent past. While aggregate domestic consumption per capita of agricultural produce, measured at 1979 prices, would be at approximately the same level in 1990 as in 1979, the sectoral shift into exports would result in decreased per capita consumption of all commodity groups except for vegetable oils (where imports would displace a large part of domestic production in total supplies) and livestock products. Measures to maintain per capita consumption of all agricultural products at 1979 levels would mean a foreign commodity trade balance for agriculture in 1990 about 20% smaller than what would otherwise result. 1.03 The crop sub-sectors are modeled in considerable detail in TASM - 46 different crop production activities can be selected in model solutions from combinations in 70 rotations, four categories of land quality are distinguished and a choice can be made between using animals and tractors as the source of draft power. The livestock sub-sector, on the other hand, is elaborated to a much lesser exteic, only seven animal production 1/ This applies to the agricultural sector only; changes in the net foreign exchange inflow into the agro-processing sector are not included. -3- activities being specified in the model. Yet the livestock sub-sector accounts for about one-third of the gross value of agricultural production in Turkey. Its contribution to sectoral value-added is higher still because much of livestock production is based on the use of relatively low cost resources such as unimproved pasture and crop residues. It is widely accepted that the scope for increasing livestock productivity in Turkey through upgrading the genetic base, better health control and improved feeding is considerable. At issue is the implication of implementing such a strategy in terms of the allocation of resources (particularly types of land) between crops for direct human use and crops for animal use and between production for export and production for domestic consumption. TASM is not well equipped to answer this because it assumes that the composition of the total feed supply to each main animal type (sheep, goat, cattle, etc.) is fixed and also that only one level of feed conversion effiency is possible with each type. Analysis of the relatively complex system of linkages between the crop and livestock sub-sectors calls for a model with less restrictive assumptions. 1.04 Modification of TASM to allow more flexible technology assumptions to be made was begun during the Turkey Agricultural Sector Study Mission's initial work with TASM and is briefly described in Report No. 4204-TU, Annex 5, Appendix 3, the new version of the model developed being referred to as TASM-ALV (Alternative Livestock Version). Further work, which is the subject of the present report, has been done since then and has resulted in another version of the model which will hereafter 'be referred to as TASM-ALV(R). (The original version of TASM as used by the Turkey Agricultural Sector Study Mission will simply be referred to here as -4- "TASM"). Model building is a task which seldom appears to be completed, the question of whether the marginal returns from further work will exceed or fall short of the marginal costs always being difficult, in prospect, to answer confidently. The conclusion here is that TASM-ALV(R) is a more useful specification than TASM: though the predictive capabilities of TASM-ALV(R) in term of the precision with which it can be made to simulate historical situations in a behaviorally plausible manner are about the same as TASM's, it provides considerably more information about the interactive components of the livestock-crop system. More work is still needed on this aspect of TASM-ALV(R), but higher priority might now be given to other modeling requirements - for example, the incorporation of internal trade between Turkey's agricultural regions and the disaggregation of production- on the basis of representative farm types and sizes. -5- II. MODELING THE LIVESTOCK SUB-SECTOR The LP Approach 2.01 The livestock sub-sector comprises an intricate set of technical and economic relationships. There are several different types of animal and for each of these there are several distinct breeds with varying productive potential as well as a spectrum of cross-breeds. Livestock husbandry systems differ from one locality to another and may be technically fairly simple or relatively sophisticated. There is a wide variety of feed sources with considerable opportunity, at certain times of the year, for substituting one source for another in response to changes in availability or cost. The decision-making process in livestock production is more complex than in crop production, with possibilities for both varying the rate of production of meat, milk, wool etc. from a given herd size and structure in the short-term and of changing this size and structure in the longer run. 2.02 In principle, the representation of technological choice in productiov at a point in time is easily done with a linear programming (LP) model -- it is a matter of specifying a sufficient number of input-output activities to reflect the segments of the production functions which are available to producers and of defining in a realistic manner the constraints on the ability of producers to mobilize resources for use in these activities. More difficult to portray are the inter-temporal trade-offs which are particularly important in the production of animals which have long reproductive cycles. All investment in TASM, including increases in livestock inventories, is exogenously specified as changes in -6- levels of resource availability. When the supply of these resources is binding in the model solution, their shadow prices provide a guide to how much loss of current consumption (in the case of livestock, reduction in herd offtake) can be tolerated in order to finance investment in increased resource availability (in the case of livestock, growth in herd size). It is theoretically possible to incorporate formal decision rules for investment in LP models like TASM but the satisfactory specification of investment activities is usually difficult and substantially increases model size and cost. It has not been attempted here and TASM-ALV(R) remains a static, one-period optimizing model. Technology Specification 2.03 In TASM, the animal types distinguished are sheep, ordinary goats, Angora goats, cattle, buffalos, mules and poultry. The production activity takes the following form: Qi = aijHj (i = 1,...,m) ...() (j = 1..n Hj = bkjXk (k = 1,...,p)(2) where Qi is the amount of the ith product produced; Hj is the number of jth type animals; Xk is the kth resource used in producing animals; aij is the amount of the ith product produced per head of the jth animal type; bkj is the amount of the kth resource used up per head in producing the jth animal type. Apart from labor, all the inputs to animals, the Xk, are types of feed and TASM distinguishes 19 different sources. These are modeled as being -7- made available as by-products of agro-processing industries (e.g. sugar beet pulp, wheat milling), as crop residues (e.g. straw), as concentrates (e.g. wheat, bar2ay, corn), or forage crops (e.g..alfalfa, cow vetch/sainfoin) and so on. Concentrates and by-products can be imported in the model. The number of animals, and hence the output of livestock products, cannot exceed the supply of feed available at the level of requirements specified by the bkj coefficients of equation (2). The main disadvantage with this formulation of livestock production is that it allows for no substitution among different types of feed. It also allows no choice of technology e.g. between low input-low output and high input-high output production systems. 2.04 Information about aggregate production functions in Turkey for the various animal types is very limited. There are no comprehensive farm survey data available to provide reliable estimates of input-output relations under different management conditions. There is substantial disagreement between the estimates from different sources of the average productivities of the national herds (output per head).1/ Estimates of the total quantities of nutrients available to livestock from feed resources also vary, though there is general agreement that while the differeice between requirements and supplies of energy over the year as a whole is fairly small there is a substantial deficiency of protein.2/ The technical coefficients specified in TASM-ALV(R) are based on farm development models 1/ See Report No. 4204-TU, Annex 5, Appendix 2, Table 8. 2/ Such calculations are, of course, dependent on the level of production taken as the benchmark for determining feed requirements and the assumptions made about feed conversion efficiency. -8- in recent Bank reports. While some of these data do reflect actual performance (mostly on Bank-assisted projects) they suffer from being drawn from small samples and from being biased towards "above-average" conditions. Details of how these data have been used for TASM-ALV(R) are given in Appendix 1. 2.05 The usual distinction made in Turkey in discussions of liurestock productivity is between production systems based on imported breeding animals with superior performance potential (for example, Brown Swiss and Holstein cattle, Merino sheep), those using genetically improved locally- bred animals and those using unimproved native animals. Given the genetically-determined potential level of performance in a herd of animals, the main factors influencing actual results (apart from selective breeding to maintain the genetic quality of the herd) will be feeding management and health control. In practice, the quality of these can vary widely and blur the differences in actual performance between herds of different potential productivity. The modeling problem is to specify production activities so as to partition this continuous spectrum of actual technical relationships into discrete performance packages which best correspond to the technology "steps" in livestock development. The number and nature of the steps identified is clearly a theoretical rather than an empirical matter and will depend on the purpose of the model. One recent Bank report, for example, distinguishes five types of technology for cattle production and four for sheep production in a project modeling exercise 1/, World Bank, Erzurum Rural Development Project -- Project File C12. another report distinguishes only three for cattle and two for sheep. 1/ A recent paper by a Turkish agricultural economist distinguished four types of technology for cattle. 2/ The choice made for TASM-ALV(R) was three types for cattle and two for sheep. The cattle activities are intended to represent (a) traditional husbandry systems using unimproved native animals fed relatively poor quality diets and receiving little or no protection from disease prevention programs; (b) systems achieving better technological performance through the use of improved cross-bred animals on a superior feeding regime and subject to active disease prevention measures; (c) intensive systems using exotic dairy cows on a high plane of nutrition under strict and effective health controls. The sheep activities specified broadly correspond to activities (a) and (b) for cattle above. No alternative technologies were formulated for goats (apart from the distinction made in TASM between ordinary and Angora types), buffalos, mules and poultry. 2.06 The specifications of these ten livestock activities in TASM-ALV(R) are given in Appendix 2, the details of how the technical coefficients were derived being given in Appendix 1. In addition to 1/ Fourth Livestock Development Project, Supervision Mission 1981: "Revised Farm Development Models." 2/ H. Gencaga, "Animal Husbandry Strategies for Turkey," paper presented to the 7th National Livestock Conference, 1981. -10- distinguishing between technology types (levels of output per head) within the total production systems for cattle and sheep, TASM-ALV(R) allows for choice in animal feeding for all types of livestock in terms of the composition of the diet, i.e. for choice of "technique" through varying factor proportions. Feed supplies are counted in energy units, but restrictions are placed on the minimum and maximum proportions of total energy intake which can be obtained from any of four groups of feed sources. The constraints are imposed to reflect physiological considerations in the formulation of livestock rations (for example, levels of roughage and dry matter) and physical limitations on producers' access to certain feeds (for example, concentrates) due to locational factors. Constraints on the protein content of feed were not specified in the model, mainly due to lack of time. This is not considered to be a serious omission because the particular categorization of feed sources into the four groups and the minima and maxima placed on the use of these will ensure that a minimum level of protein is available for the more advanced technologies. Thus the livestock production activities in TASM-ALV(R)are of the following form: Qi Z 3 aijAj (i = 1,...,m) .(3) (j = 1,...,10) A < cjEj ... (4a) A < djL *..(4b) Ej =Z grjFr (r = 1,2,3,4) ...(5) (0 K minrj grj N maxrj' 1) ( Z ri =1) Fr = hrsxs (s = 1,...,20) ...(6) s -11- where Qi is the amount of the ith product produced (meat, milk, wool or draft power); A is the number of animals produced with the jth type of technology (e.g. "traditional" sheep, "intensive" cattle, etc.); E is the amount of energy available for consumption by animals produced with the jth type of technology; L is the amount of labor available for agriculture; Fr is the rth group of feed sources (pasture - including meadow hay, concentrates, straw and crop residues, forage crops); Xs is the amount of the sth individual feed source (e.g. sugar beet pulp, wheat straw, alfalfa hay) available; aij is the amount of the ith product produced per head of animals produced with jth type technology; c is the number of animals produced with jth type technology per unit of energy available for consumption by these animals; dj is the number of animals produced with jth type technology per unit of labor used; 9rj is the proportion of total energy available for consumption by animals produced with jth type technology supplied from the rth group of feed sources; hrs is the amount of energy provided per unit of the sth individual feed source in the rth group; Within the cereals component of the F2 group of feed sources (concentrates) minimum and maximum proportions of the contribution of different cereals to -12- total energy available from cereals are also specified (see Appendix 2 for details). Sensitivity tests, reported in Appendix 3, show that model solutions in terms of relative levels of A- are fairly sensitive to values of Qij/Ej (product/energy ratios of the different technologies), grj (minimum/maximum proportions of energy allowed from the various types of feed), and the crop composition of the cereal component in total feed consumed. Setting Resource Availabilities 2.07 Feed supplies. The total area of land available for agricultural use in the base year (1979) is assumed to be 49.9 million ha of which 25 million ha are assumed to be pastures and meadows. The corresponding values used for TASM are 49.9 million ha and 21.7 million ha respectively. Both versions of the model thus adopt a figure for pasture and meadows higher than TOPRAKSU's estimate for the early 1970s and considerably higher than the estimate obtained from the SIS 1980 Census of Agriculture. 1/ As noted in the Turkey Agricultural Sector Study Mission's report, the SIS source is of uncertain reliability and much of the area reported as being "unused" must be in some form of agricultural use. 2/ It is assumed here that much of this area comprises marginal land fit only for extensive grazing and that some of this land so used has been categorized as "forest and scrub" in the published SIS statistics on land use. 3/ The balance of the agricultural land area (24.9 million ha) is available for allocation among crops in TASM-ALV(R) solutions. The structures of land resource 1/ Report No. 4204-TU, Annex 5, Appendix 2, Table 3. 2/ Report No. 4204-TU, Annex 1. 3/ SIS, Agricultural Structure and Production, 1980. -13- availability assumed for TASM and TASM-ALV(R) are compared in Appendix 4, Table A4.1 together with simulations of land use in various model runs. 2.08 Yields per ha of crops, crop by-products and crop residues which can be fed to livestock are specified in the crop production activities of the model. Crop processing activities also yield feed supplies, and most types of crops and all types of concentrates produced domestically can appear as imported supplies in the model solutions. These physical quantities are converted into energy terms using nutritive values obtained from standard tables. Since the latter were prepared for use in developed countries the validity of their application to Turkish feed sources and feeding systems is, to a certain extent, questionable. A greater area of uncertainty, however, is the average value for ruminant nutrition to be ascribed to Turkey's entire pasture and meadow land.1 2.09 Labor supplies. The figure for maximum manpower available for agricultural work, i.e. 3.1 million hours per quarter in 1979, assumed in TASM is also used in TASM-ALV(R). 2.10 Livestock inventories. Total numbers in 1979 for the seven types of livestock specified in TASM (i.e. sheep, cattle, ordinary goats, Angora goats, buffalos, mules and poultry) are used in TASM-ALV(R). For the latter version of the model, sheep and cattle are divided among the different technologies as follows: 1/ The figure adopted for TASM-ALV(R) assumes that 1 ha of this land can provide 55% of the annual energy needs of one head of a native cattle herd reared in the traditional manner. -14- Assumed Distribution of Sheep and Cattle Production Technologies in 1979 Livestock husbandry Number of animals Proportion system (technology type) ('000) (%) Sheep -'traditional' 43,725 95.0 -'improved' 2,301 5.0 TOTAL 46,026 100.0 Cattle -'traditional' 13,232 85.0 -'improved' 2,257 14.5 -'intensive' 78 0.5 TOTAL 15,567 100.0 These figures are clearly little more than guesstimates of the situation and are only meaningful when read in conjunction with the input-output coefficients assumed for each technology (see Appendix 2). Some support for this view of production conditions in the case of cattle.is provided by Gencaga (1981) who cites estimates of 'traditional' cattle farms as accounting for over 85% of the total herd (though half of this categor' is said to be subject to "strict veterinary control"), "improved traditional" farms as accounting for nearly 13% and "modern" farms for under 2%. 1/ Little confirmatory information is available for sheep production, but the proportion of 'improved' animals, at least, is known to be small. 1/ Op-cit, Table 1. -15- Model Validation 1/ 2.11 The levels of commodity production, consumption, international trade and prices adopted for used in TASM as the benchmark data to describe the actual situation in 1979 were carried over into TASM-ALV(R) with the exception of figures for livestock products. The TASM production levels for meat and milk are based on SPO estimates and are considered by many to be very much on the high side. The TASM-ALV(R) figures reflect judgment that actual meat and milk production in Turkey is much more likely to correspond to the levels of per capita consumption revealed by the 1981 household consumption survey data reviewed by the Turkey Agricultural Sector Mission.2/ The contrasting assumptions made for the two versions of the model are shown below: Assumed Production Levels for Livestock Products in 1979 ('000 mt) TASM TASM-ALV(R) Meat: sheep 338.0 230.6 goat 110.0 77.6 cattle 391.0 229.3 buffalo 34.0 19.8 poultry 132.0 132.0 TOTAL 1,005.0 689.3 Milk: sheep 1,102.2 614.4 goat 626.0 372.3 cattle 3,386.4 2,913.1 buffalo 296.6 218.5 TOTAL 5,411.2 4,118.3 Eggs: poultry ('000) 4,322.7 4,322.7 1/ All solutions reported in this paper are solved with the official foreign exchange rate of TL47/US$. 2/ See Report No. 4204-TU, Annex 3, Table 1. -16- Assumed wool production differs very little between the two versions. However, TASM-ALV(R) does not specify any production of hides, whereas TASM does. Output of animal draft power is not changed. Since both versions of the model assume the same herd sizes for the base year, the implied productivities are very different: Production per Head in 1979 (kg) ( no. per ) Meat Milk Eggs (1000 head) TASM TASM-ALV(R) TASM TASM-ALV(R) TASM TASM-ALV(R) Cattle 25.12 14.67 /a 217.50 160.10 /a Sheep 7.34 4.96 /a 24.02 13.20 /a Ordinary goat 6.85 4.70 38.32 21.00 Angora goat 1.77 1.80 15.00 15.00 Buffalo 32.68 19.00 285.10 210.00 Poultry 2.24 2.24 76.37 76.37 /a Weighted averages of the different technologies. Consumption levels (differences between domestic production and net foreign trade) in TASM-ALV(R) for livestock products are adjusted downwards from TASM values, but prices are transferred unchanged.1/ 2.12 The differences between TASM and TASM-ALV(R) can be summarized as follows: (a) base year actual values for the production and consumption of livestock products are a quarter to a third lower in. TASM-ALV(R) than in TASM; 1, In theory, price elasticities of demand should be re-estimated at the new levels of consumption. However, previous work with TASM had shown that model solutions are insensitive to these elasticity values (see Report No. 4204-TU, Annex 5, para. 10) and so no changes to them were made for TASM-ALV(R). -17- (b) the TASM-ALV(R) solution is optimized over three activities for cattle production and two for sheep production compared with only one activity for each animal .type in TASM. (c) the TASM-ALV(R) solution represents the least-cost combination of feed sources for any given level of output of the livestock sector. This is not necessarily the case with the TASM solution. 2.13 The base solutions for agricultural production of the two versions of the model are compared in Table 1. Although the fit of total production to the actual 1979 value is closer with TASM-ALV(R), this version is wide of the actual values for commodity groups/sub-sectors by a greater margin than TASM in every case. TASM-ALV(R) under-predicts livestock output by nearly 10%. In the TASM solution, herd size is a binding constraint on all types of animals, but this is not the case in the TASM-ALV(R) solution as shown in Table 2 1/. On the other hand, pasture availability is a binding constraint in the TASM-ALV(R) solution whereas in TASM some pasture remains unused despite the 'maximum' possible number of livestock being carried. Othar aspects of resource use given by the base 1 It is interesting to note that the base solution values for numbers of sheep, goats and cattle are closer to the (unpublished) estimates of the SIS Agricultural Census for 1980, viz ('000) Census 1980 TASM-ALV(R) 1979 Sheep 41,510 41,332 Goats 16,045 14,300 Cattle 12,196 13,505 -18- solutions are also shown in Table 3. TASM-ALV(R) uses less arable land than TASM and also less labor and fertilizer, but a very similar amount of tractor power. Shadow prices for livestock products in the base solutions for the two versions of the model are compared in Table 4. The under-prediction of production levels by TASM-ALV(R) is reflected in higher shadow prices than those given by TASM. 2.14 A breakdown of the total amount of feed energy consumed by each type of livestock technology in the TASM-ALV(R) base solution into the four groups of feed sources is shown in Table 5. It can be seen that the two cheapest sources of feed, pasture (Group A) and straws and crop residues (Group B) tend to be used up to the maximum allowed by the constraints on feed composition spdcified. In contrast, consumption of the relatively expensive forage crops (Group D) is kept to the minimum. Table 5 also shows that the proportion of barley in total cereals used for livestock feed is in the middle of the permissible range 65-75%. -19- III. POLICY EXPERIMENTS AND PROJECTIONS Objectives 3.01 As was done with TASM, a number of runs were made with TASM-ALV(R) to determine the patterns of production, consumption and international trade for the agricultural sector which are optimal under alternative market structures, states of technology and levels of resource availability to those observed (or assumed) for the base year, 1979. Two sets of simulations were undertaken. (a) a set with the same states of technology and resource availabilities assumed as those in the base solution but with different market structures, the purpose being to see how the agricultural sector would have looked under a different trade policy from the one which actually prevailed in 1979; (b) a set with different states of technology, resource availabilities and market structures from those used in the base solution and representing conditions which could prevail in 1990, the purpose being to see how the agricultural sector would look in such circumstances. 3.02 As with TASM, the reason for simulating counterfactual behavior in the base year is to be able to identify the direction, and gauge the order of magnitude, of responses to alternative trade policies. When these same policies are also assumed in simulations for projection years the. contribution to growth between the base and projection year of the policy -20- change can then be distinguished from the contribution of other exogenously-specified changes intended to reflect such factors as shifts in technology, increased domestic demand due to population and per capita income growth and investment in higher levels of resource availability. Base period simulations which assume different trade policies from those which actually prevailed in 1979 are only meaningful as an analytical tool for investigating the role of policy in determining agricultural production, consumption and trade patterns in a situation projected into the future. Thus, trade policy experiments set in the context of historical situations only cannot be interpreted as predictions of what would happen if the counterfactual policies were to be implemented. Any simulation which implicitly or explicitly represents movement forward in time from the base year requires that values for all exogenous variables which are likely to change with time be re-specified, not just the values of the policy instruments. Policy Experiments under Base Year Conditions 3.03 The Turkey Agricultural Sector Study Mission was concerned, among other things, with assessing the impact on the agricultural sector of continuing liberalization of government policy on international trade. Various experiments were run on TASM to assist this analysis. The one repeated on TASM-ALV(R) is that designated as "Policy III". This represents a foreign trade regime in which there are no restrictions on the quantities of agricultural produce which can be imported into Turkey and no tariff barriers. Agricultural exports are similarly free of taxes (or subsidies) but are assumed to face a limited level of foreign demand at -21- base period CIF prices. In the case of grains, it is assumed that marginal revenue from exports in excess of 2.5 million mt declines by one-third.,/ 3.04 A comparison of the solutions of the two versions of the model for Policy III under base period conditions for production, consumption and international trade is shown in Table 6. The most striking feature is the lower production and consumption levels for wheat in the TASM-ALV(R) solution (exports are identical). 2/ Table 7 shows that the inroads into domestic consumption caused by the greater sectoral emphasis on exports are considerably greater with TASM-ALV(R) than with TASM. Thus total consumption is more than 3% below the base solution compared with a slight increase above the base solution in the TASM result. Grains and livestock consumption is 11% and 4%, respectively, below base period levels in TASM-ALV(R) compared with only 1% below and 3% above, respectively, in TASM. Comparisons of the Policy III base period solutions for resource use are given in Table 8. The availability of irrigated land and the main category of rainfed land is a binding constraint on the solutions for both versions of the model, but pasture supply is only binding in TASM-ALV(R). 3.05 Table 9 compares the Policy III solution of TASM-ALV(R) for livestock output and feed consumption with the base solution of the same version of the model. Numbers of all types of animals are higher under Policy III conditions, but technology shifts within the sheep and cattle groups are in opposite directions. This is reflected in the-different 1/ The limitations specified on export volumes in the model are shown in Appendix 4, Table 2. 2/ Most of the difference in solution values for the livestock sector shown in Table 6 are due to the higher levels assumed for actual base period figures in TASM compared with TASM-ALV(R). See para. 2.11. -22- composition (by energy values) of the feed consumed under Policy III from the composition in the base solution for 'traditional' sheep and both 'traditional' and 'improved' cattle. For example, greater competition from the greater profitabilityl/ of the 'traditional' cattle under Policy III forces the 'improved' cattle to maximize consumption of the low cost Group A feed (pasture) and reduce consumption of the higher cost Group B feeds (concentrates). 'Traditional' cattle have done the opposite: minimizing Group A intake and increasing Group B, indicating that for this type of technology it is profitable to 'pay' more for more nutritive feed, given the efficiency of energy conversion assumed (see Appendix 3 for details of other experiments affecting livestock-feed economics). It will be noted from Table 9 that the composition of the cereal component of Group B feed consumption is the same under Policy III as in the base solution. Projections 3.06 The projection procedures and assumptions used for TASM-ALV(R) are identical to those for TASM 2/ with the single exception of the way in which technological change in the livestock subsector is modeled. In TASM, the input-output coefficients for all crop and livestock production activities are multiplied by selected values to give new sets of factor-product and factor-factor ratios. In TASM-ALV(R), the same procedure and same multipliers are used for all activities except the sheep and cattle activities. In these cases, it is not the technology which is changed but the proportions of the different types of technology represented within the animal groups. Thus the improvement of productivity 1/ Profitability in economic terms (i.e. relative to real resource costs). 2/ Described in Report No. 4204-TU, Annex 5. -23- in the sheep and cattle sub-sectors overall is seen as the result of the progressive recruitment of producers and their animals into improved systems of animal husbandry. The productivity of 'traditional' systems, for example, is not assumed to change over time, but it is assumed that by 1990 a smaller proportion of the national herd/flock will be reared in this way than in 1979. The amount of change in the technology composition of cattle and sheep production which could be projected was limited by setting maximum constraints on livestock numbers as follows: TASM-ALV(R): Assumed Maximum Possible Changes in the 'Technology Mix' of Sheep and Cattle Production, 1979 to 1990 1979 1990 ('000) % ('000) Sheep -'traditional' 43,725 95.0 38,898 80.0 -'improved' 2,301 5.0 9,724 20.0 TOTAL 46,026 100.0 48,622 100.0 Cattle -'traditional' 13,232 85.0 10,031 61.0 -'improved' 2,257 14.5 6,249 38.0 -'intensive' 78 0.5 165 1.0 TOTAL 15,567 100.0 16,445 100.0 The maximum increase in total herd/flock size allowed was 5.6%, the same as in the TASM projections. The technology composition with maximum livestock numbers in 1990 was set so as to give overall increases in the productivity of sheep and cattle production which would be close to those assumed in TASM: -24- Changes in Sheep and Cattle Productivities, 1979-1990 in TASM and TASM-ALV(R) With Policy III (kg/head) 1979 1990 TASM TASM-ALV(R) TASM TASM-ALV(R) Sheep: Meat 7.34 5.01 8.07 (+10.0%)a/ 5.64 (+12.6%) Milk 24.02 13.35 26.42 (+10.0%) 14.40 (+7.9%) Cattle: Meat 25.12 14.73 30.14 (+20.0%) 17.45 (+18.5%) Milk 217.50 187.13 261.00 (+20.0%) 229.12 (+22.4%) a/ Figures in brackets are the changes between 1979 and 1990. Subject to the constraints on the technology composition of cattle and sheep production imposed by maxima set on livestock numbers produced by the different technologies, productivity change within cattle and sheep production is projected endogenously by TASM-ALV(R). In TASM, productivity change is purely an exogenous assumption. As it happens, the Policy III projections of TASM-ALV(R) showed livestock numbers to be binding constraints on the solution, so the productivity change projected represented the maximum allowed. 3.07. TASM and TASM-ALV(R) projections to 1990 for production, consumption and international trade under Policy III are summarized in Tables 10 and 11. There is not a great deal of difference betweet the total sectoral and sub-sectoral growth rates and the sub-sectoral -25- composition of total production projected by the two versions. 1/ The major difference occurs within the grain sub-sector: TASM-ALV(R) projects an absolute decline in wheat production at an implied average rate of nearly 2% p.a. whereas TASM projects a modest expansion at 0.8% p.a. Consequently, TASM-ALV(R) does not project any exports of wheat in 1990 in contrast to the US$313 million worth (2.5 million mt) projected by TASM. The fixed input proportions technology assumptions in TASM's livestock activities mean that a relatively large quantity of wheat straw is always required per unit of livestock output (see Appendix 2, Table A2.1) and in the TASM 1990 solution wheat straw does have a positive shadow price. In TASM-ALV(R), on the other hand, other cereal and pulse straws can be substituted for wheat straw which diminishes its relative value.2/ Wheat is thus more likely to be displaced by other crops such as barley in the solution. This linkage between wheat and livestock production was explored further through some additional experiments which are discussed briefly in Appendix 3. 3.08 Details of projected livestock feed consumption and production are given in Tables 12 and 13. Apart from Angora goats, all lives.ock inventories are used to their maximum which represents a 5.6% increase above 1979 levels. This additional pressure on the pasture resource (Group A in Table 12) ensures that the cereal and pulse straws (Group C) continue 1/ The difference between the shares of the livestock sub-sector projected by TASM and TASM-ALV(R) mainly reflect the different assumptions made about these shares for the base year. 2/ TASM-ALV(R) only takes into account the nutritive contribution of crops to livestock production. Substantial amounts of-crop material (mostly straw) are also needed for winter bedding. However, all the TASM-ALV(R) solutions generate more than enough cereal straw to meet this need over and above the maximum which could be used for feeding. -26- to be used to the limit allowed. The projection of total energy intake obtained from pasture decreases for 'traditional' sheep, ordinary goats, 'traditional' cattle and buffalo, all of which correspondingly increase the proportion of their energy intake obtained from concentrates (Group B). As in the base solutions, consumption of forage crops is kept to the minimum possible. 3.09 The TASM-ALV(R) projection solution uses less crop land than TASM but more pasture (see Table 14). Tractor draft power and fertilizer use is lower in TASM-ALV(R) but the labor requirement is substantially higher. As Table 15 shows, while both versions of the model project similar absolute levels of labor input per ha for 1990, the increase over the base solution values is 23-25% (the range reflecting alternative ways of measuring employment - see Table 15 for explanation) in the case of TASM-ALV(R) compared with only 9-13% in the case of TASM. Conversely, the increase in tractor power input per ha between 1979 and 1990 is greater in the TASM solutions than in the TASM-ALV(R) solutions. Projected levels of, and implied rates of change in, fertilizer use per ha are very similar in both solutions. The growth rates in agricultural employment and labor productivity between 1979 and 1990 implied by the model solutions are shown in Table 16. TASM-ALV(R) depicts employment growing as fast as value-added and considerably in excess of the rate of growth assumed for the agricultural labor force (1% p.a.), in contrast to TASM which projects an increasing, not a decreasing, degree of agricultural (i.e. rural) 'unemployment'. 3.10 Part of the difference between the results of the two versions of the model is accounted for by greater selection in the TASM-ALV(R) solution of animal-biased technology in the crop activities as opposed to -27- tractor-biased technology.1/ However, most of the difference is explained by the contrasting assumptions made in the two versions of the model regarding changes in labor productivity in livestock production. In TASM, no adjustments are made to the product output/labor input coefficients (measured as man-hours/anf"mal/year) in livestock production activities in transferring from base year simulations to projection year simulations. Thus labor productivity in livestock production is projected to increase by 10% between 1979 and 1990 in the case of sheep and goat production, by 20% in the case of cattle and buffaloes (see para. 3.06) and by 71% in the case of poultry and not to change at all in the case of mules. In TASM--ALV(R), however, no rate of increase in labor productivity is assumed a priori in the case of cattle and sheep production -- the outcome depends on the projections of the different types of technology within these groups which are selected in the solution. The increase in the sheep and cattle populations in the TASM projections of 5.6% over the base year level requires the equivalent of an.extra 86,000 man-years of labor (i.e. 1.5-2.1% on top of the base year employment level) whereas the increase in herd/flock size in TASM-ALV(R) of 7.7% requires an extra 329,000 man-years (i.e. 6.1-8.4% on the base year level). Thus virtually no increase in the number of animals per worker is projected for cattle and sheep production in TASM-ALV(R). This is not unreasonable since the better feeding practices, stricter health controls and general upgrading of stockmanship 1/ Nearly all the crop production activities in the model allow for the draft power input to be provided by animals or tractors, depending on which is most profitable, with appropriate adjustments to the technical coefficient. In both TASM and TASM-ALV(R) projections with Policy III, animal power availability is a binding constraint on the solutions. -28- required to achieve greater productivity per animal are all labor-demanding improvements, 3.11 A breakdown of the total projected changes in sectoral production, consumption, international trade and employment between 1979 and 1990 into a component reflecting the impact of Policy III and a component reflecting the impact of growth assumptions (changes in domestic demand at constant prices, technology and resources availabilities) is shown in Tables 17 and 18. The relative contributions of the 'trade policy' and 'growth' effects to the difference in production levels between base and projection year solutions is very similar for both versions of the model -- a little over one-third due to trade policy and a little under two-thirds due to growth. TASM and TASM-ALV(R) solutions are also in fairly close agreement about the relative importance of the two components in accounting for changes in exports and imports. The two versions of the model differ, however, in their results with regard to domestic consumption changes. In the projected situation, TASM-ALV(R) indicates a quite strong negative impact from Policy III on consumption whereas TASM indicates almost no impact at all. As the discussion in para. 3.10 above would suggest, TASM and TASM-ALV(R) also give quite contrary results for the sources of change in employment and labor productivity (Table 18). Thus the growth factor works against employment according to TASM, but for employment according to TASM-ALV(R). -29- IV. CONCLUSIONS 4.01 Making the link between crop and livestock production less rigid in TASM by allowing feed consumption costs to be maximized through substitution among feed energy sources radically changes the model's solutions for the grains subsector. The key issue is whether the assumptions now fncorporated in TASM-ALV(R) are more or less realistic than those built into TASM. The scope which exists in practice for combining feedstuffs in different proportions will depend largely on the extent to which livestock breeding and fattening can be geographically separated and the amount of inter-regional trade in feedstuffs which is feasible. Disaggregation of TASM-ALV(R) into agricultural regions 1/ with inter-regional competition allowed would be the logical next step. 4.02 Livestock production technology is quite difficult to describe satisfactorily in the model, in the sense of being able to produce in the base solution the combination of different technologies within cattle, sheep and goat production assumed to occur in reality. This is a familiar problem in LP models when continuous non-linear functions are represented by linear segments. Relatively small changes in the input-output coefficients can cause some activities to disappear entirely in the solution, resulting in very angular production functions. For TASM-ALV(R), the answer is probably to disaggregate livestock production by type and size of Gnterprise. Together with the suggestion made in para. 4.01, this implies a major investment in further model development. Thus the 1/ Work has already begun on a regional version of TASM but this has been confined so far to disaggregation of the resource base and production activities. No disaggregation of internal market structures and formulation of an inter-regional trade matrix has been undertaken. -30- conclusion is that TASM-ALV(R) probably represents the most which it is sensible to do in terms of elaborating the structures of livestock production at the national level. 4.03 The main insight which TASM-ALV(R) has added to the assessment of the medium term prospects for Turkish agriculture afforded through the use of TASM is that the potential for an expansion of barley production relative to wheat production appears to be even higher, and therefore a reduction in wheat consumption. On the other hand, the rural "unemploy- ment" is declining d"e to higher demand of labor in the livestock sub- sector. The greater profitability of barley in drier areas combined with a strong demand both directly in the export market for coarse grains and indirectly in the export market for livestock products will put it in a strong economic position in the future. Wheat, on the other hand, will be more responsive than barley to movements in the prices of grains relative to the prices of other agricultural commodities in exporL markets unless wheat production is maintained through the existing system of input subsidy and output price support and restrictions on international trade. -31- Table 1: COMPARISON OF TASM AND TASM-ALV(R) BASE SOLUTIONS FOR GROSS VALUE OF AGRICULTURAL PRODUCTION IN 1979 A B C D 1979 1979 1979 Actual TASM TASM-ALV(R) C/A ($M) a/ ($M) ($M) (%) Grains 2,311 2,277 2,172 94.0 of which wheat 1,483 1,502 1,427 96.2 others 828 775 745 90.0 Pulses 311 349 352 113.2 Vegetables 2,580 2,851 2,855 110.7 Fruits and nuts 2,607 2,745 2,745 105.3 Oil crops 480 499 508 105.8 Industrial crops 1,149 1,153 1,106 96.3 Livestock products 2,718 (3,824)b/ 3,827 2,463 90.6 TOTAL 12,156 (13,262)b/ 13,701 12,201 100.4 a/ All values at actual 1979 farmgate prices. b/ Figures in brackets are assumed actual 1979 values for livestock and total agricultural production, respectively, in TASM. -32- Table 2: COMPARISON OF TASM AND TASM-ALV(R) BASE SOLUTIONS FOR LIVESTOCK PRODUCTION A B C D 1979 1979 1979 C/A PRODUCTION ('000 mt) Actual a/ TASM b/ TASM-ALV(R) ( Sheep - Meat 230.6 338 202.7 87.9 Milk 614.4 1,105.5 544.4 88.6 Wool 40.8 59.4 36.5 89.4 Goat - Ordinary - Meat 71.0 103.5 55.1 77.7 Milk 317.3 579 246.4 77.7 Wool 9.1 9.1 7 77.4 - Angora - Meat 6.6 4.7 4.6 70 Milk 55 40 38.5 70 Wool 5.8 4.2 4.1 69.9 Cattle - Meat 229.3 391 197.3 86 Milk 2,913.1 3,385.8 2,505.3 86 Buffalo - Meat 19.8 34 19.8 99.8 Milk 218.5 296.6 218.5 100 Poultry - Meat 132 132 132 100 Eggs ('000) 4,322.7 4,501.1 4,501.1 104 HERD COMPOSITION ('000 head) c/ Sheep - 'traditional' 43,725 46026* 40,318 92.2 'improved' 2,301 1,014 44.1 Goat - Ordinary 15,109 15,109* 11,733 77.7 Angora 3,666 2,667 2,567 70 Cattle - 'traditional' 13,232 11,643 88 'improved' 2,257 15,567* 1,784 79 'intensive' 78 78* 100 Buffalo 1,040 1,040* 1,040* 100 Poultry 58,939 58,939* 58,939* 100 Mule, etc. 2,453 2,453* 2,453* 100 OUTPUT PER HERD (kg) Sheep - Meat 4.73 7.34 4.9 103.6 Milk 13.3 24.02 13.17 99 Wool 0.9 1.29 0.88 97.8 Goat - Ordinary - Meat 4.7 6.85 4.7 100 Milk 20.7 38.32 21 101.4 Wool 0.6 0.6 0.6 100 Angora - Meat 1.8 1.77 1.8 100 Milk 15 15 15 100 Wool 1.4 1.58 1.58 112.9 -33- Table 2/cont. Cattle - Meat 14.68 25.12 14.61 99.5 Milk 187.1 217.5 185.5 99.1 Buffalo - Meat 19 32.68 19 100 Milk 210 285.1 210 100 Poultry - Meat 2.24 2.24 2.24 100 Eggs (no./1000 hd) 73.3 76.37 76.37 104.2 a/ World Bank estimates b/ Livestock production and output per head in TASM is calibrated to higher base period estimates than TASM-ALV(R). See para. 2.11. c/ Totals for animal types are SIS published figures. The breakdowns of numbers within total cattle and sheep are World Bank estimates. * Indicates that the number of livestock shown is the maximum allowed by the livestock inventory constraint specified in the model. -34- Table 3: COMPARISON OF TASM AND TASM-ALV(R) BASE SOLUTIONS FOR INPUT USE IN 1979 1979 1979 TASM TASM-ALV(R) Land ('000 ha) .Cultivated Area a/ 19,012 17,788 of which: sown 12,586 11,644 fallow 6,426 6,144 Under tree crops 2,279 2,279 Pasture 20,377 25,000* Labor ('000 persons) b/ 5,617 5,345 Tractors ('000 units) c/ 67 66 Fertilizer ('000 mt) d/ Nitrogen 792 745 Phosphate 817 775 a! Net cultivated area, i.e. multiple cropped land is counted only once. b/ Numbers employed during the peak quarter of the year in terms of male adult full-time equivalents. It is assumed that six hours a day are spent actually in the fields working and 300 days a year are worked. c/ Number of units required during the peak quarter of the year. It is assumed that each unit supplies 1500 hours of work a year. d/ Quantities of nutrient required. * Indicates that the availability of this resource is a binding constraint on the model solution. -35- Table 4: COMPARISON OF TASM AND TASM-ALV(R) BASE SOLUTIONS FOR PRICES OF LIVESTOCK PRODUCTS IN 1979 (US$/mt) Farmgate Import a/ Export a/ (actual) (actual) (actual)- TASM TASM-ALV(R) Sheep - Meat 1,210.64 2,220.0 786.92 1,523.16 Milk 378.94 382.72 519.14 Wool 3,605.96 4,315.80 4,315.80 5,565.02 Goat - Ordinary - Meat 962.98 2,220.0 972.61 1,492.62 Milk 265.96 268.62 460.11 Wool 2,112.34 700.2 1,753.24 5,601.64 - Angora - Meat 1,008.51 2,220.0 1,563.19 1,744.72 Milk 265.96 507.98 666.19 Wool 5,720.00 804.6 6,082.26 7,836.40 Cattle - Meat 1,321.92 1,855.3 1,335.13 1,573.08 Milk 304.26 307.30 390.30 Buffalo - Meat 1,286.38 1,855.3 1,140.00 1,299.25 Milk 272.55 275.28 275.28 Poultry - Meat 3,436.27 762.0 2,233.51 2,233.51 Eggs 70.21 70.92 70.92 a/ Parity values at farmgate. -36- Table 5: TASM-ALV(R) BASE SOLUTION FOR LIVESTOCK ENERGY CONSUMPTION IN 1979 Energy Energy Group Composition Required (% of total energy) a/ Per Head b/ A B C D. Sheep - 'traditional' 110 50 Mc/ 25 25 M 0 'improved' 180 35 M 30 15 M 20 m Goat - Ordinary 127 50 M 25 25 M 0 Angora 175 35 40 M 25 M 0 Cattle - 'traditional' 360 27.2 37.8 35 M 0 'improved' 550 25 55 M 15 M 5 m 'intensive' 1,200 15 M 45 0 40 m Buffalo 580 35 M 30 35 M 0 Poultry 2.55 0 100 m 0 0 Mule, etc. 417 20 m 45 35 M 0 Group B - Cereals Composition '000 mt % Wheat 436.9 10 Corn 655.3 15 Rye 218.4 5 Barley 3,057.0 70 a/ Energy can be obtained from four groups of feed sources: A - pasture/meadow it (grazing/hay); B - concentrates (cereals, processing by-products and oil seed cakes); C - straws and crop residues; D - forage crops (alfalfa, sainfoin/ vetch). b/ In kg of Starch Equivalent per year (World Bank estimates) c/ M (m) denote maximum (minimum) proportions by energy value allowed in the total feed mix. Table 6: COMPARISON OF TASM AND TASM-ALV(R) SIMULATIONS OF PRODUCTION, CONSUMPTION AND INTERNATIONAL TRADE UNDER BASE PERIOD CONDITIONS WITH POLICY III a/ (US$ million) b/ PRODUCTION CONSUMPTION c/ IMPORTS EXPORTS TASM TASM TASM TASM TASM -ALV(R) TASM -ALV(R) TASM -ALV(R) TASM -ALV(R) Grains 2,733 2,494 1,600 1,391 0 0 784 806 of which - wheat 1,638 1,410 1,279 1,079 0 0 314 314 - others 1,095 1,084 321 312 .0 0 470 492 Pulses 556 559 272 275 0 0 308 308 Vegetables 3,060 3,065 2,811 2,816 0 0 225 225 Fruits and nuts 3,025 3,025 2,313 2,313 0 0 206 206 Oil crops 422 431 433 444 13 15 4 4 Industrial crops 1,423 1,388 821 785 0 0 732 732 Livestock products 3,887 d/ 2,521 3,434 2,309 12 22 806 377 TOTAL 15,106 13,483 11,684 10,333 25 37 3,065 2,658 a/ No policy restrictions on international trade, but the marginal revenue from grains exports declines once a certain volume has been reached and the absorptive capacity of export markets at base period CIF prices is limited in the case of other commodities. b/ Production and consumption are valued at actual 1979 farmgate prices; international trade at actual 1979 import/export farmgate parity prices. c/ Consumption is defined as all uses (including seed, waste and loss) apart from animal feed and exports. .d/, The actual 1979 level of livestock production assumed is 41% higher in TASM than in TASM-ALV(R). -38- Table 7: COMPARISON OF TASM AND TASM-ALV(R) SIMULATIONS OF EFFECTS OF POLICY III a/.ON PRODUCTION AND CONSUMPTION UNDER BASE PERIOD CONDITIONS (% change from the respective base solutions) PRODUCTION CONSUMPTION b/ TASM TASM TASM -ALV(R) TASM -ALV(R) Grains 20.1 14.8 -0.8 -11.1 of which - wheat 9.1 -1.2 -0.1 -12.6 - others 41.3 45.5 -3.0 -5.7 Pulses 59.4 58.8 -2.8 -2.8 Vegetables 7.3 7.4 0.2 0 Fruits and nuts 10.2 10.2 -1.0 -1.0 Oil crops -15.4 -15.2 6.4 7.0 Industrial crops 23.5 25.5 -6.9 -7.3 Livestock products 0 2.3 3.1 -4.3 TOTAL 9.8 10.5 0.3 -3.2 a/ No policy restrictions on international trade but the marginal revenue from grain exports declines once a certain volume has been reached and the absorptive capacity of export markets at base period CIF prices is limited in the case of other commodities. b/ Consumption is defined as all uses (including seed, waste and loss) apart from animal feed and exports. -39- Table 8: COMPARISON OF TASM AND TASM-ALV(R) SIMULATIONS OF INPUT USE UNDER BASE PERIOD CONDITIONS WITH POLICY III a/ TASM TASM-ALV(R) Land ('000 ha) Cultivated area b/ 24,014 21,741 of which: sown 15,791 14,670 fallow 8,223 7,071 Under tree crops 2,369 2,369 Pasture 20,359 25,000* Labor ('000 persons) c/ 6,219 6,094 Tractors ('000 units) d/ 67 89 Fertilizer ('000 mt) e/ Nitrogen 940 885 Phosphate 987 911 a/ No policy restrictions on international trade but the marginal revenue from grain exports declines once a certain volume has been reached and the absorptive capacity of export markets at base period CIF prices is limited in the case of other commodities. b/ Net cultivated area, i.e. multiple cropped land is only counted once. c/ Labor figures represent the number of male adult full- time equivalents employed during the peak quarter of the year. It is assumed that six hours a day are spent actually in the fields working and 300 days a year are worked. d/ Tractor figures represent the number of units required during the peak quarter of the year. Assumed annual work rate is 1,500 hrs/unit. e/ Quantity of nutrient required. * Indicates that the availability of this resource is a binding constraint on the model solution. Table 9: TASM-A LV(R): SIMULATION OF LIVESTOCK PRODUCTION AND ENERGY CONSUMPTION UNDER BASE PERIOD CONDITIONS WITH POLICY III a/ Energy Consumption by Feed Source (%) Livestock Numbers ('000) Group A Group B Group C Group D Base Solution Policy III BS P-III BS P-III BS P-III BS P-III Sheep - 'traditional' 40,318 42,304 50 M1 b/ 50 M 25 15 25 M 25 H 0 10 'improved' 1,014 2,301* 35 m 35 M 30 30 15 M 15 M 20 m 20 m Goats - Ordinary 11,733 12,789 50 M 50 M 25 25 25 M 25 M 0 0 Angora 2,567 2,567 35 35 40 M 40 M 25 M 25 M 0 0 Cattle - 'traditional' 11,643 13,232* 27.2 20 M 37.8 45 35 M 35 M 0 0 'improved' 1,784 1,002 25 35 M 55 M 45 15 M 15 H 5 m 5 m 'intensive' 78 78* 15 M 15 M 45 45 0 M 0 M 40 m 40 m Buffalo 1,040* 1,040* 35 M 35 M 30 30 35 M 35 M 0 0 Mule, etc. 2,453* 2,453* 20 m 20 m 45 45 35 M 35 M 0 0 Poultry 58,939* 58,939* 0 0 100 M 100 M 0 0 0 0 Composition of Group B Cereals (% by weight) BS P-III Wheat 10 10 Corn 15 15 Rye 5 5 Barley 70 70 a/ No policy restrictions on international trade but the marginal revenue from grain exports declines once a dertain volume has been reached and the absorptive capacity of export markets at base period CIF prices is limited in the case of other commodities. b/ M (m) denote maximum .(minimum) proportions by energy value allowed in the total feed mix. * Indicates that the iiumber of livestock shown is at the maximum allowed by the livestock inventory constraint specified in the model. -41- Table 10: COMPARISON OF TASM AND TASM-ALV(R) PROJECTIONS OF OuTPUT'AND INTERNATIONAL TRADE IN 1990 WITH POLICY III a/ A B C D LIMITED EXPORTS, UNLIMITED RESTRICTED TRADE IMPORTS (POLICY III) Base Solution Projections GROWTH RATE 1979-1990 1979 1990 TASM TASM-ALV(R) TASM TASM-ALV(R) TASM TASM-ALV(R) A-C B-D PRODUCTION ------- ($M at 1979 actual iarmgate prices) -------(% p.a.)------ Total 13,701 12,201 17,610 15,530 2.3 2.2 Grains 2,227 2,172 2,349 2,532 1.7 1.4 of which: wheat 1,502 1,427 1,632 1,157 0.8 -1.9 others 775 745 1,117 1,375 3.4 5.7 Pulses 349 352 497 497 3.3 3.2 Vegetables 2,851 2,855 3,557 3,672 2.0 2.3 Fruit and nuts 2,745 2,745 3,644 3,620 2.6 2.5 Oil crops 499 508 349 349 -3.2 -3.4 Industrial crops 1,153 1,106 1,676 1,662 3.5 3.8 Livestock products 3,827 2,463 5,138 3,198 2.7 2.4 CONSUMPTION ------- ($M at 1979 actual farmgate prices) ------- ------(% p.a.)------- Total 11,654 10,676 14,691 12,651 2.1 1.6 Grains 1,612 1,565 1,597 1,408 -0.1 -1.0 of which: wheat 1,281 1,234 1,259 1,064 -0.2 -1.3 others 331 331 338 344 0.2 0.4 Pulses 280 283 332 331 1.6 1.4 Vegetables 2,805 2,816 3,363 3,478 1.7 1.9 Fruit and nuts 2,336 2,337 2,745 2,747 1.5 1.5 Oil crops 407 415 544 555 2.7 2.7 Industrial crops 882 847 1,051 1,036 1.6 1.8 Livestock products 3,332 2,413 5,059 3,096 3.9 2.3 EXPORTS ($M at 1979 actual farmgate export parity prices) ------( p.a.)------- Total 765.6 776.8 2,342.1 1,935.2 10.7 8.7 Grains 156.0 156.0 671.9 352.8 14.2 7.7 of which: wheat 154.6 154.6 312.6 0 6.6 - others 1.4 1.4 359.3 352.8 65.6 65.3 Pulses 72.5 72.5 192.2 192.6 9.3 9.3 Vegetables 14.7 14.7 181.7 181.7 25.7 25.7 Fruit and nuts . 43.7 43.7 326.1 311.2 20.1 19.5 Oil crops 4.7 4.7 0 0 - - Industrial crops 396.5 396.5 731.9 731.9 5.7 5,7 Livestock products 77.5 88.7 238.3 165.1 12.5 5.8 IMPORTS ($M at 1979 actual farmgate import parity prices) -----(% p.a.)------- Total 9.8 25.9 280.5 309.5 35.7 25.3 Grains - 0 0 0 0 - - Pulses 0 0 0 0 - - Vegetables 0 0 0 0 - - Fruit and nuts 0 0 0 0 - - Oil crops 0 0 147.1 156.5 - - Industrial crops 0 0 0 0 - - Livestock products 9.8 25.9 133.4 153.0 26.8 17.5 a/ No policy restrictions on international trade but the marginal revenue from grain exports declines once a certain volume has been reached and the absorptive capacity of export markets at base period CIF prices is limited in the case of other commodities. Table 11: COMPARISON OF TASM AND TASM-ALV(R) PROJECTIONS OF PRODUCTION AND EXPORT COMMODITY COMPOSITION IN 1990 WITR POLICY III a/ (%) PRODUCTION EXPORTS PROJECTION (1990) PROJECTION (1990) BASE SOLUTION (1979) b/ POLICY III BASE SOLUTION (1979) POLICY III TASM TASM TASM TASM TASM -ALV(R) TASM -ALV(R) TASM -ALV(R) TASM -ALV(R) Grains 16.6 17.8 15.6 16.3 20.4 20.1 28.7 18.2 of which: wheat 11.0 11.7 9.3 7.4 20.2 19.9 13.3 0 others 5.6 6.1 6.3 8.9 0.2 0.2 15.4 18.2 Pulses 2.5 2.9 2.8 3.2 9.5 9.3 8.2 10.0 Vegetables 20.8 23.4 20.2 23.6 1.9 1.9 7.8 9.4 Fruit and nuts 20.0 22.5 20.7 23.3 5.7 5.6 13.9 16.1 Oil crops 3.6 4.2 2.0 2.2 0.6 0.6 0 0 Industrial crops 8.4 9.1 9.5 10.7 51.8 51.0 31.3 37.8 Livestock products 27.9 20.2 29.2 20.6 10.1 11.4 10.2 8.5 TOTAL c/ 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 a/ No policy restrictions on international trade but the marginal revenue from grain exports declines once a certain volume has been reached and the absorptive capacity o; export markets at base period CIF prices is limited in the case of other commodities. b/ Base solution commodity compositions in TASM and TASM-ALV(R) differ substantially because of the higher weight given to livestock products in the actual 1979 data for the former model. c/ Percentages may not sum exactly to 100.0 due to rounding. Table 12: TASM-ALV(R): PROJECTIONS OF LIVESTOCK NUMBERS AND ENERGY CONSUMPTION IN 1990 WITH POLICY III a/ Livestock Numbers ('000) Energy Consumption by Feed Source (%) Base Solution Policy III Group A Group B Group C Group D 1979 1990 BS P-III BS P-III BS P-III BS P-III Sheep - 'traditional' 40,318 38,898* 50 M b/ 35 25 40 M 25 M 25 M 0 0 'improved' 1,014 9,724* 35 H 35 M 30 30 15 M 15 M 20 m 20 m Goats - Ordinary 11,733 15,961* 50 M 35 25 40 M 25 M 25 K 0 0 Angora 2,567 1,835 35 35 40 M 40 M 25 M 25 H 0 0 Cattle - 'traditional' 11,643 10,031* 27.2 20 m 37.8 45 35 M 35 H 0 0 'improved' 1,784 6,249* 25 28.6 55 M 51.4 15 M 15 M 5 m 5 m 'intensive' 78* 165* 15 H 15 M 45 45 0 M 0 H 40 m 40 m Buffalo 1,040* 1,099* 35 M 20 m 30 45 35 H 35 M 0 0 Mule, etc. 2,453* 2,453* 20 m 20 m 45 45 35 M 35 M 0 0 Poultry 58,939* 62,263* 0 0 100 M 100 m 0 0 0 0 Composition of Group B Cereals (% by weight) BS P-II Wheat 10 10 Corn 15 15 Rye 5 5 Barley 70 70 a/ No policy restrictions on inter"ational trade but the marginal revenue from grain exports declines once a certain volume has been reached and the absorptive capacity of export markets at base period CIF prices is limited in the case of other commodities. b/ M (m) denote maximum (minimum) proportions by energy value allowed in the total feed mix. * Indicates that the number of livestock shown is at the maximum allowed by the livestock inventory constraint specified in the model. -44- Table 13: COMPARISON OF TASM AND TASM-ALV(R) PROJECTIONS OF LIVESTOCK PRODUCTION AND INTERNATIONAL TRADE IN 1990 WITH POLICY III a/ LEVEL OF PROPORTION OF LEVEL OF PRODUCTION b/ INTERNATIONAL TRADE PRODUCTION TRADED TASM TASM TASM TASM -ALV(R) TASM -ALV(R) TASM -ALV(R) --------------- ('000 mt) ----------------- ----- (%) ------ Sheep - Meat 392.6 274.2 2.6 c/ 22.8 0.7 8.3 Milk 1,284.7 700.2 Goat - Ordinary - Meat 120.3 75.0 26.2 11.5 21.8 15.3 Milk 672.8 335.2 - Angora - Meat 7.5 3.3 0.6 0 8.0 0 Milk 63.9 27.5 Cattle - Meat 495.7 287.0 80.6 46.3 16.3 16.1 Milk 4,292.1 3,767.9 Buffalo - Meat 43.1 20.9 7.6 1.7 17.6 8.1 Milk 376.0 230.8 Poultry - Meat 238.5 139.5 12.2 0 5.1 0 Eggs d/ 8,131.1 4,755.0 a/ No policy restrictions on international trade but the marginal revenue from grain exports declines once a certain volume has been reached and the absorptive capacity of export markets at base period CIF prices is limited in the case of other commodities. b/ The base level of livestock production (actual 1979) assumed is 41% higher in TASM than in TASM-ALV(R). c/ Positive numbers denote exports, negative numbers denote imports. d/ '000. -45- Table 14: COMPARISON OF TASM AND TASM-ALV(R) PROJECTIONS OF INPUT USE IN 1990 WITH POLICY III-a/ TASM TASM-ALV(R) Land ('000 ha) .Cultivated Area b/ 16,950 15,322 of which: sown 11,151 10,244 fallow 5,799 5,078 Under tree crops 2,183 2,177 Pasture 20,409 25,000* Labor ('000 persons) c/ 6,021 6,396 Tractors ('000 units) d/ 107 97 Fertilzer ('000 mt) e/ Nitrogen 1,395 1,292 Phosphate 1,911 1,303 a/ No policy restrictions on international trade but the marginal revenue from grain exports declines once a certain volume has been reached and the absorptive capacity of export markets at base period CIF prices is limited in the case of other commodities. b/ Net cultivated area, i.e. multiple cropped land is only counted once. c/ Labor figures represent the number of male adult full time equivalents employed during the'peak quarter of the year. It is assumed that six hours a day are spent actually in the fields working and 300 days a year are worked. d/ Tractor figures represent the number of units required during the peak quarter of the year. Assumed annual work rate is 1,500 hrs/unit. e/ Quantity of nutrient required. * indicates that this level of resource is binding on the model solution. -46- Table 15: COMPARISON OF TASM AND TASM-ALV(R) PROJECTIONS OF FACTOR USE INTENSITIES IN 1990 WITH POLICY III a/ Base Solution Projection 1979 1990 TASM TASM-ALV(R) TASM TASM-ALV(R) LABOR b/ man-days per ha: I 40.4(100)c/ 35.6(100) 45.7(113) 44.4(125) II 30.1(100)- 26.2(100) 32.8(109) 32.2(123) TRACTOR POWER d/ tractor-hours per ha: I 5.3(100) 5.6(100) 9.5(79) 9.5(170) II 3.5(100) 3.7(100) 6.9(197) 6.3(170) mechanized area (%):e/ 36.5(100) 35.8(100) 74.8(205) 68.8(192) FERTILIZER (kg/ha)f/ 75.6(100) 75.7(100) 146.7(194) 148.3(196) IRRIGATION irrigated area (%)5/ 14.7(100) 15.7(100) 20.9(142) 23.1(147) a/ No policy restrictions on international trade but the marginal revenue from grain exports declines once a certain volume has been reached and the absorptive capacity of export markets at base period CIF prices is limited in the case of other commodities. b/ Man-days per total used agricultural land (net cultivated area plus pasture). Male adult equivalents, assuming 6 hours per day actually spent in the fields and 300 days worked per year. I = Number on the basis of hours of employment during the peak quarter of the year. II = Number on the basis of total hours of employment during all quarter of the year. c/ Figures in brackets are indices with 1979 values = 100. d/ Tractor-hours per net cultivated ha, excluding the area under tree ^'rops. I = tractor-hours required during the peak quarter of the year multiplied by four; II = sum of tractor hours required during each- quarter of the year. e/ Proportion of net cultivated area, excluding the-area under tree crops, accounted for by mechanized crop activities selected in the model solution. f/ Quantities of nitrogen and phosphate nutrient per net cultivated ha. gi Proportion of net cultivated area, excluding the area under tree crops. 14 -47- Table 16: COMPARISON OF TASM AND TASM-ALV(R) PROJECTIONS OF EMPLOYMENT AND VALUE-ADDED IN 1990 WITH POLICY III a/ (Monetary values at actual 1979 farmgate prices) LIMITED EXPORTS, UNLIMITED IMPORTS RESTRICTED TRADE (POLICY III) Base Solution Projection 1979 1990 TASM TASM-ALV(R) TASM TASM-ALV(R) Gross value of production - Total ($M) 13,701 12,201 17,610 15,530 - Growth rate 1979-90 (% p.a.) 2.31 2.22 Value-added in agriculture - Total ($M) 11,914 10,491 14,161 12,491 - Growth rate 1979-90 (% p.a.) 1.58 1.60 Number employed in agriculture b/ ( 5,617 5,345 6,021 6,395 - Total ('000) I ( 0.63 1.64 - Growth rate 1979-90 (% p.a.) ( 4,173 3,931 4,321 4,629 II ( 0.32 1.50 Value-added per worker c/ ( 2,121 1,963 2,352 1,953 - Total ($) I( 0.94 -0.05 - Growth rate 1979-90 (% p.a.) ( 2,855 2,669 3,277 2,698 II ( 1.26 0.10 Labor content of total costs (%) d/ 69 69 54 59 Labor content of net income (%) d/ 33 36 29 35 a/ No policy restrictions on international trade but the marginal revenue from grain exports declines once a certain volume has been reached and the absorptive capacity of export markets at base period CIF prices is limited in the case of other commodities. b/ Male adult equivalents, assuming 6 hours per day actually spent in the fields and 300 days worked per year. I = Number on the basis of hours of employment during the peak quarter of the year. II = Number on the basis of total hours of employment during all quarters of the year. c/ 'Value-added' = gross value of production less cost of seed, fertilizer, animal/tra:ctor power and certain other working capital items. Thus the-costs of fixed asset investment and other overheads are not taken into account. Values at 1979 actual farmgate prices. d/ All labor, whether hired or supplied to the household by the household, is costed at TL25/hour. Table 17: COMPARISON OF TASM AND TASM-ALV(R) PROJECTIONS WITH POLICY III a/ I . Comparison of Policy Simulations for Base Year (1979) and Projection Year (1990) (US$ million) b/ A B C RESTRICTED TRADE LIMITED EXPORTS LIMITED EXPORTS UNLIMITED IMPORTS UNLIMITED IMPORTS (POLICY III) (POLICY III) i ii iii Base Solution Projection B - A C - B C- - A 1979 1979 1990 TASM TASM TASH TASM TASM TASM TASM -ALV(R) TASM -ALV(R) TASM -ALV(R) TASM -ALV(R) TASM -ALV(R) TASM -ALV(R) Production 13,701 12,201 15,106 13,483 17,610 15,530 1,405 1,282 2,504 2,047 3,909 3,329 Consumption 11,654 10,676 11,684 10,333 14,691 12,651 30 -343 3,007 2,318 3,037 1,975 Exports 766 777 3,065 2,658 2,342 1,935 2,299 1,881 -723 -723 1,576 1,158 Imports 10 26 25 37 281 309 15 11 256 272 271 -283 00 II . Sources of Change 1979-90 (Z) Trade Policy Effect Growth Effect i / iii ii / iii TASM TASM-ALV(R) TASM TASM-ALV(R) Production 36 39 64 61 Consumption 1 -17 99 117 Exports 146 162 -46 -62 Imports 6 4 94 96 a/ No policy restrictions on international trade but the marginal revenue from grain exports declines once a certain volume has been reached and the absorptive capacity of export markets at base period CIF prices is limited in the case of other commodittes. b/ All values at 1979 farmgate prices (import or export parity where appropriate). Table 18: COMPARISON OF TASM AND TASK-ALV(R) PROJECTIONS OF EMPLOYMENT AND LABOR PRODUCTIVITY WITH POLICY III a/ I . Comparison of Policy Simulations for Base Year (1979) and Projection Year (1990) A B C RESTRICTED TRADE LIMITED EXPORTS LIMITED EXPORTS UNLIMITED IMPORTS UNLIMITED IMPORTS (POLICY III) (POLICY III) i ii iii Base Solution Projection B - A C - B C - A 1979 1979 1990 TASM TASM TASM TASM TASM TASK TASM -ALV(R) TASM -ALV(R) TASM -ALV(R) TASM -ALV(R) TASM -ALV(R) TASM -ALV(R) No. employed in agriculture ('000) b/ I: 5,617 5,345 6,219 6,094 6,021 6,395 602 749 -198 301 404 1,050 II: 4,173 3,931 4,469 4,348 4,321 4,629 296 417 -148 281 148 698 Value added per worker ($) I: 2,121 1,963 2,043 1,850 2,352 1,953 -78 -113 309 103 231 -10 II: 2,855 2,669 2,843 2,593 3,277 2,698 -12 -76 434 105 422 29 I II . Sources of Change 1979-90 (Z) Trade Policy Effect Growth Effect i / iii ii / iii TASM TASM-ALV(R) TASM TASM-ALV(R) Number employed in agriculture ('000) 1: 149 71 -49 29 II: 200 60 -100 40 Value-added per worker ($) I: -34 -1130 134 1030 II: -3 -262 103 362 a/ No policy restrictions on international trade but the marginal revenue from grain exports declines once a certain volume has been reached and the absorptive capacity of export markets at base period CIF prices is limited in the case of'other commodities. b/ Male adult equivalents, assuming 6 hours per day actually spent in the fields and 300 days worked per year. I -Number on the basis of hours of employment during the peak quarter of the year. II - Number on the basis of total hours of employment during all quarters of the year. -50- Appendix 1: TECHNICAL COEFFICIENTS FOR LIVESTOCK PRODUCTION ACTIVITIES IN TASM-ALV(R) Method of Calculation 1.01 In the TASM-ALV(R) livestock production activities, output of physical product for human consumption (meat, milk and wool) is in fixed proportion to the total number of animals produced by the respective type of technology. Coefficients for the base year (1979) are obtained by dividing estimates of total production by estimates of livestock populations. For TASM-ALV(R), the former data are constructed from estimates of consumption per capita at the retail level obtained from recent household surveysl/ while tUe latter are read directly from SIS published statistics.2/ The estimates of total production so obtained are considerably lower than, for example, SPO estimates but the coefficients for animal productivity derived from these lower values agree well with Bank estimates of average productivity based on field observation and cited in project reports. 1.02 In TASM-ALV(R), the inputs of labor and feed energy are also in fixed proportion to the total number of animals produced by each type of technology. The labor coefficients are the same as in TASM except in the case of the 'improved' sheep, 'intermediate' cattle and 'intensive' cattle activities where man-hour requirement per animal head was increased above the TASM value (details are given in Appendix 2). The feed energy coefficients were calculated as follows: (a) Feeding requirements for three types of cattle production (native, dairy-beef and imported dairy) and two types of sheep 1/ See Report No. 4204-TU, Annexes 3 and 5. 2/ SIS, Agricultural Structure and Production, 1980. The data are for animal numbers at the end of the year. -51- production (pre-project and with-project for Mor-karaman animals) in terms of quantities of straw, hay, barley etc. for different components of the herds such as lactating cows, yearling sheep etc. were taken from the report of the Supervision Mission of the Fourth Livestock Development Project (LDP4)1/. These data are shown in Table A1.1. (b) These data were then converted to an annual basis by multiplying by the appropriate factor. For example, if 500 kg of meadow hay are required by each Animal Unit for 180 days, then the annual energy requirement is equivalent to 500 x (365/180) - 1,000 kg.2/ (c) Herd projections given in the same report, the particular components to which the feed requirements are addressed, the liveweight of marketable animals and the marketable quantity of milk produced (i.e. net of consumption by young stock) were calculated per 100 head of the animal populations recorded at the end of the year in the SIS statistics. These estimates are shown in Tables A1.2 and A1.3. (d) The values obtained from stages (b) and (c) were used to calculate total annual feed requirements per head of livestock. These physical quantities for different types of feed were then expressed in terms of a common energy unit -- the Starch Equivalent (SE) - using factors for kg SE per kg feed derived 1/ Turkey, Fourth Livestock Development Project, Supervision Mission October 18 - November 13, 1981: "Revised Farm Development Models". 2/ For the case where the requirement for the particular feed would be of less than full duration, as for lactation or for young animals which would be culled or sold before passing into the next annual cohort, then a different numerator from 365 applies. -52- from data in reference tables of nutritive values1/. (The factors used are shown in Appendix 2). Thus the production function in the livestock activities in TASM-ALV(R) is expressed in terms of kg of output of meat, milk or wool per kg of input of SE. These feed conversion coefficients for cattle and sheep and corresponding indexes are shown in Table A1.5. The amount of energy required from grazing was calculated as the difference between total enery requirements and the amount of energy supplied by the specific quantities of feed specified in Table A1.1. Four main groups of feed were then distinguished on the basis of the physiological needs of animals for different nutritive factors (energy, protein, dry matter content digestibility), as follows: Group A: Pasture (intake from grazing and consumption of meadow hay); Group B: Concentrates (whole cereals, crop by-products from processing including oilseed cakes and sugar beet pulp); Group C: Roughage (cereal and pulse straw); Group D: Forage Crops (alfalfa, vetch and sainfoin hay). The composition of total feed intake from these groups for the production systems specified in the LDP4 report is shown in Table A1.4. If P. McDonald et al, Animal Nutrition, Oliver & Boyd, 1973, 2nd Edition. The Starch Equivalent (SE) and Total Digestible Nutrients (TDN) systems of feed energy accounting have been superseded by systems in which the energy values of feeds are-expressed in terms of metabolizable energy. The SE system was used here because of the writers' familiarity with it and because it is perfectly adequate for the purposes of TASM-ALV(R). -53- (e) The final step involved adjustment of the feed conversion and productivity per head coefficients derived from the LDP4 data in order to arrive at technology specifications which were more representative of general conditions in the Turkish livestock sub-sector. The LDP4 data are based on "the actual performance observed in the better project farms (emphasis added) in the project area of Livestock IV and in the eastern part of the project area of Livestock III."1/ After some review of other information on livestock production in Turkey and some experimentation with different values (see Appendix 3) , the coefficients labelled "TASM-ALV(R) assumptions" in Table A1.5 were adopted. (f) Once the coefficients for feed energy input per head had been determined for the sheep and cattle production activities, those for other activities (ordinary goat, Angora goat, buffalo and mule) were adjusted in proportion to the relative energy requirements implied by feed inputs in the TASM activities (Appendix 2). The coefficient for poultry was kept the same as in TASM. Evaluation 1.03 The three most critical aspects of livestock energy conversion specification for a one-period optimizing model such as TASM are giving the correct expression to: (a) the total pressure on the feed resource base of the livestock population of a given size and structure; 1/ Turkey op. cit. para. 1. -54- (b) the relative demands for energy of the populations of different animal type (sheep, cattle, goat, etc.); (c) the relative demands for energy from different technologies within a population of a particular animal type ('traditional' vs. 'improved', etc.). Independent checks on (a) are very difficult to obtain mainly because of the poor information concerning the amount and quality of livestock feed availability in Turkey due to uncertainties about actual land use and cropping patterns. The plausibility of the ratios of type (b) are easier to gauge because the relative nutritional needs of different types of livestock are fairly well known though most data relate to performance under more intensive management than generally prevails in Turkey. The relative feed energy requirements assumed in TASM-ALV(R) for the different types of livestock appear quite reasonable on this basis, though it might be argued that the ratio of cattle to sheep requirements is perhaps not high enough (see the indexes in Table A1.5). Since the ratios of type (c) are derived largely from hypothetical constructs in the absence of hard information about the distribution of technologies in the livestock sub-sector, their validity is difficult to assess, except in terms of other models. In Figure A1.1, the three pairs (one for meat, one for milk) of feed energy conversion ratios for the 'traditional', 'improved' and 'intensive' technologies asumed in TASM-ALV(R) are plotted with the corresponding feed energy conversion ratios calculated from the LDP4 data. -55- It can be seen that the TASM-ALV(R) 'production function' is assumed to retain the same basic configuration as that derived from the LDP4 data 1/. A similar graphing of the corresponding data for sheep (Figure A1.2) shows that a higher feed energy conversion efficiency is assumed in TASM-ALV(R) than is implied in the analysis of the LDP4 data. 1/ The cross-over of the beef and milk lines in Figure A1.1 reflects the fact that the most intensive feeding levels are found in the specialized dairy cattle technologies. - Table 41.1 FEEDING REQUIREMENTS FOR CATTLE AND SHEEP al Native Cattle Dairy-Beef Cattle Exotic Dairy Cattle Per Animal Unit b/ Per Lactating Cow Per Animal Unit c/ Per Lactating Cow Per Yearling Per Calf Per Animal Unit d/ Per Lactating Cow Per Yearling Per Calf M days days k days daydays days days g days days kg days Maintenance Straw 1,000 180 360 265 Meadow Hay 500 180 1,080 265 600 365 Legume Hay 360 265 2,400 365 Growth Straw Meadow Hay Legume flay Concentrate (factory six) 190 365 V4 365 274 365 Barley 126 365 Production Straw Meadow flay I-gume flay 1,000 el 365 1.000 E/ 365 Concentrate (factory mix) 192 120 I/ 560 f/ 365 Barley 400 e/ 365 Sheep 'Pre-Proj ect' Sheep 'With Project' Per Eve Per Ram Per Yearling Per Lamb Per Ewe Per Ram Per Yearling Per Lamb FZ Days !1 Days Days k Days g Days g Days Kg Days g Days Ul Mainotenance Straw 120 190 102 160 52.5 160 2.1 30 4eadow Hay 238 190 238 160 122.5 160 4.9 30 190 190 190 160 120 160 5 30 Legume flay 190 190 190 160 120 160 5 30 Growth & Production Barley 25 365 6 365 33 365 9 365 &/ For each herd compopent, the amount of feed required (which includes an allowance of 10% for storage loss and wastage) is shown in the left-hand column (in kg) and the period of time corresponding to this feed input I shown in the right-hand column (in days). Feed during the remainder of the year (if any) is obtained from grazing. b/ Mature cow: 0.5 AU 1-2 year old female: 0.3 AU 2-3 year old female: 0.4 AU Bull: 0.6 AU 1-2 year old male: 0.3 AU Calf: 0.1 AU (assumed for the purposes of this report) cl Mature cow: 1.0 AU 1-2 year old female: 0.6 AU 2-3 year old female: 0.8 AU Breeding Kale: 1.2 AU 1-2 year old male: 0.6 AU 6-12 month old calf: 0.3 AU' 0-6 month old calf: 0.15 AU (assumed for the purposes of this report). d/ Same as for dairy-beef cattle. e/ For 2,500 kg milk per lactation. f/ For 2:900 kg milk perjacration. k/ Lactation time is 180 days in total. SOURCE: Turkey. Fourth Livestock Development Project - Supervision Mission, October/November 1981. "Revised Farm Development Models', Appendix III. -57- APPENDIX 1 Table A1.2 HERD STRUCTURES AND PRODUCTION FOR CATTLE Number per head at year-end Animal Lactating Units a/ Cows Yearlings b/ Calves c/ Native cattle 46.7 37.3 20.0 31.8 Dairy-beef cattle 93.1 40.9 23.5 44.0 Exotic dairy cattle 99.3 55.6 21.3 42.3 a/ Including calves. b/ No. of yearlings = no. of 1-yr. olds less half the disappearance of 1-2 yrs. olds. c/ no. of calves = no. born less half the disappearance of calves. Liveweight production (kg per head at year-end) CULLS SALES Native cattle 17.88 15.46 Dairy-beef cattle 40.52 60.24 Exotic dairy cattle 63.39 50.72 Source: Based on data in Turkey, Fourth Livestock Development Project - Supervision Mission, October/November 1981. "Revised Farm Development Models", Appendixes I and II. -58- APPENDIX 1 Table A1.3 HERD STRUCTURES AND PRODUCTION FOR SHEEPS Number per head at year-end Ewes Rams Yearlings Lambs Sheep 'pre-project' 63.0 2.4 25.6 40.6 Sheep 'with-project' 55.6 2.2 28.3 46.1 Liveweight production (kg per head at year-end) CULLS SALES Sheep 'pre-project' 4.02 6.73 Sheep 'with-project' 6.40 14.72 Source: Based on data in Turkey, Fourth Livestock Development Project - Supervision Mission, October/November 1981. "Revised Farm Development Models", Appendixes I and II. APPENDIX 1 Table A1.4 COMPOSITION OF ENERGY INTAKE a/ BY CATTLE AND SHEEP Dairy Beef Exotic Dairy Sheep Sheep Native Cattle Cattle Cattle Pr-Proje' 'With Project' kg/hd % kg/hd kg/hd % )cg/hd Z kFg/hd Z Feeds Straw 60.7 16.9 43.6 3.9 10.4 5.4 Meadow Hay 93.4 26.1 402.2 36.2 238.3 15.1 74.7 38.6 58.5 23.0 Legume Hay 260.5 23.5 1,028.7 65.0 51.2 20.1 Concentrate (factory mix) 46.5 13.0 54.3 4.9 315.6 19.9 Barley 137.2 12.4 12.9 6.7 15.9 6.2 Grazing b/ 157.6 44.0 213.3 19.1 95.7 49.4 128.7 50.6 TOTAL 358.2 100.0 1,110.1 100.0 1,582.6 100.0 193.7 100.0 254.3 100.0 TASM-ALV(R) Feed Groups A (roughage: pasture/meadow hay) 70 55 15 88 74 B (concentrate: cereals, by-products, oilseed cakes) 13 17 20 - - 7 6 C (straws: cereal, pulse) 17 4 0 5 0 D (forage: alfalfa, vetch, sainfoin) 0 24 65 0 20 a/ 'Energy is measured in units of Starch Equivalent. b/ Difference between total energy requirements as calculated from Tables 1, 2 and 3 and the energy contributed by the other feeds listed above. 1i v y APPENDIX 1 Table A1.5 ENERGY/PRODUCTION RELATIONS FOR CATTLE AND SHEEP Cattle Sheep 'LOw' 'Intermediate' 'High' 'Lower' 'Higher' Performance Performance Performance Performance Performance Energy Consumption (kg SE/head/year) (a) LDP 4: 385 1,110 1,583 194 254 (b) 'Theoretical' 990 185 (c) TASN-ALV(R) assumption 360 550 1,200 110 180 Meat Production (kg/head/year) (a) LDP 4: .15.0 50.3 59.3 5.1 11.0 (c) TASH-ALV(R) assumption: 13.0 24.0 40.0 4.8 9.0 Milk Production (kg/head/year) (a) LDP 4: 233 807 1,470 17.6 24.4 (c) TASM-ALV(R) assumption: 160 325 800 13.0 20.0 INDEXES OF ENERGY REQUIREMENTS (kg SE/head/year) Cattle Sheep Cattle Sheep 'Low' 'Intermediate' 'High' 'Lower' 'Higher' 'Intermediate' 'Iigher' Performance Performance Performance Performance Performance Performance Performance . (a) LDP 4 185 573 817 100 131 436 100 (b) 'Theoretical' 534 100 (c) TASM-ALV(R) assumption 327 500 1,091 100 164 305 1 INDEXES OF ENERGY CONVERSION EFFICIENCY (kg output/kg SE input) Cattle Sheep Cattle Low 'intermediate' 'High' 'Lower' 'Higher' 'Low' 'Intermediate 'High' Performance Performance Performance Performance Performance Performance Performance Performance Heat (a) LDP 4 149 174 144 100 166 100 117 97 (b) TASM-ALV(R) assumption 83 100 76 100 115 100 121 92 Milk (a) LDP 4 666 800 1,022 100 106 100 120 153 (b) TASM-ALV(R) assumption 376 500 564 100 94 __W 133 150 Notes: LDP 4: From Appendices I, II and III. 'Theoreticallri Calculatd by applying feeding standarda to hard cospositto and production data in Tables 1 *n4 2. -61- Flgure AI.1: INPUT-OUTPUT RELATIONS FOR CATTLE PRODUCION OUTFRT (Kg/heod) Meot/Milk 80/1400 - 60/1200 / e' 50/1000 - 40/800 - 30/600- 20/400 10/200 I I i I I I I I I i I I i I I J 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600• --- ---DrD FOM DATA IN LIESTOCK IV SE (Kg/hood) INRJT SLERVISON M~SSON REPOfT ASSUMPONS USE0 IN TASM-ALV (R) Wedld Bank-24804 -62- Figure A1.2: INPUT-OUTPUT RELATIONS FOR SHEEP PRODUCTION OUTRT (Kg/head) Meat/Milk 15/30 10/20- 5/10- 100 200 300 SE ftha)NU ** - - --DERIVED FROM DATA IN UVESTOCK IV S K/ed NU SUPERVISION MISSION REPORT m ASSUMPTIONS USED IN TASM-ALV (R) World 80nk-24805 -63- Appendix 2: TASM-ALV(R) LIVESTOCK SUB-SECTOR DATA 2.01 TASM input requirements for the livestock sub-sector are much more restrictive than the requirements specified in TASM-ALV(R). The input coefficients in TASM are shown in Table A2.1.A. They are pasture, labor and 16 different grains, straws and concentrates. All these are given in fixed proportions with no substitution allowed between pasture and/or any feed. In terms of Starch Equivalents, the sources of energy by feed group per animal type are shown in Table A2.1.B. 2.02 The input-output coefficients for TASM-ALV(R) livestock sub-sector are shown in Table A2.2. Also in this table are the minimum and maximum proportions that each feed group can contribute, in terms of constraints specified in the model, towards the total energy required by each animal technology. Comparison of Tables A2.1 and A2.2 indicates that the two systems are very similar. The difference is that TASM-ALV(R) distinguishes two breeds for sheep and three breeds for cattle. In addition, TASM-ALV(R) allows for a certain amount of substitution among feed groups. The advantage of these modifications is that the trade-offs between feeding systems can be analyzed. 2.03 Table A2.3 shows the energy value (in Starch Equivalents) of the various sources of feed and Table A2.4 shows the minima and maxima that each cereal type can contribute towards the total cereals consumed by animals. The constraints on model solutions shown in Table A2.4 are . derived from Turkish historical data. -64- APPENDIX 2 Table A2.1 INPUT COEFFICIENTS IN TASM LIVESTOCK ACTIVITIES (for comparison with the input coefficients in TASM-ALV(R)) A . Annual Requirements Sheep Goat Angora Cattle Buffalo Mule Poultry Pasture (ha/head) 0.17 0.17 0.17 0.5 0.5 0.5 Labor (man-hrs/head) 11.53 11.53 11.53 142 65 78 .66 Feed a/ (kg/head) Wheat 4.6 5.0 7.4 18.8 26.2 18.8 .4 Corn 4.6 5.0 7.4 19.0 26.4 19.0 .2 Rye 1.9 2.1 3.2 8.1 11.2 10.7 .1 Barley 21.4 23.3 34.4 88.1 122.6 88.4 1.1 Fodder 0.6 0.7 0.7 2.6 3.5 2.9 F-Wheat 106.0 111.7 123.9 432.5 587.9 484.4 F-Corn 16.5 17.4 19.3 67.3 91.5 75.4 F-Rye 5.8 6.2 6.8 23.9 32.5 26.7 F-Barley 36.2 38.2 42.2 147.6 200.7 165.3 F-Pulses 5 5.2 5.8 20.4 27.7 22.9 F-Alfalfa 7.1 7.4 8.3 28.8 39.2 32.3 F-Fodder 3.2 3.4 3.8 13.1 17.8 14.7 C-Wheat 11.4 12.4 18.3 46.9 65.3 47.1 .6 C-Rye .3 .3 .4 1.1 1.6 1.1 .01 C-Barley .05 .05 .08 .2 .3 .2 .004 C-Sugar beet 30.2 32.8 48.5 124.2 172.8 124.5 1.6 B . Annual Requirements in Terms of Sources of Feed Energy (Kg SE/ per head) Feed Group c/ A 34.0 34.0 34.0 100.0 100.0 100.0 0 B 47.3 51.4 76.0 194.5 270.7 196.7 2.5 C 26.5 28.0 31.0 108.2 147.1 121.2 0 D 5.8 6.1 6.7 23.5 31.9 26.3 0 Total 113.6 119.5 147.7 426.2 549.7 444.2 2.5 a/ F- stands for crop by-products and C- for processing by-products. b/ Starch Equivalents. C/ Feed Groups as specified in TASM-ALV(R). APPENDIX 2 Table A2.2 INPUT-OUTPTT COEFFICIENTS IN TASM-ALV(R) LIVESTOCK ACTIVITIES AND ALLOWABLE PROPROTIONS IN FEED COMPOSITION Sheep Goat Cattle Buffalo Mule Poultry 'Traditional' 'Improved' Ordinary Angora 'Traditional' 'Improved' 'Intensive' INPUTS AND OUTPUTS Labor (man-hrs/head) 11.53 13.84 11.53 11.53 142 170.4 284 65 78 0.66 Energy Required (kg SE/head) 110 180 127 175 360 550 1,200 580 417 2.55 Animal Power (hrs/head) 47 60 120 Meat (kg/head) 4.8 9 4.7 1.8 13 24 40 19 2.24 Milk (kg/head) 13 20 21 15 160 325 800 210 Wool (kg/head) 0.88 1 0.6 1.58 Eggs (no./1,000 heads) 76.37 In FEED GROUP (in percentage, by energy contribution) a/ LEVELS Minimum-A 30 20 30 30 20 10 - 20 20 - Maximum-A 50 35 50 50 35 35 15 25 35 - Minimum-B 5 10 5 5 10 15 20 10 10 100 Maximum B 40 50 40 40 50 55 60 50 50 100 Minimum-C 10 - 10 10 10 5 - 10 10 - Maximum-C 25 - 15 25 25 35 15 - 35 35 - Minimum-D - 20 - - - 5 40 - - Maximum-D 80 80 80 80 80 80 80 80 80 - a/ For example, the amount of' energy which Feed Group B can supply in the production of 'traditional' cattle ranges from a minimum of 10% to a maximum of 50%-of'the.total energy supplied from all feed sources. -66- APPENDIX 2 Table A2.3 ENERGY VALUE (STARCH EQUIVALENT) PER UNIT OF INPUT Group-A Group-B Group-C Group-D (kg SEa//ha) Pasture 200 (kg SE/kg) b/ Wheat 0.72 Corn 0.78 Rye 0.65 Barley 0.71 C-Wheat 0.50 C-Rye 0.24 C-Barley 0.60 C-Sugar beet 0.60 K-Sunflower 0.53 K-Groundnut 0.56 K-Cotton 0.54 K-Soybean 0.68 F-Wheat 0.13 F-Corn 0.15 F-Rye 0.17 F-Barley 0.23 F-Pulses 0.19 F-Alfalfa 0.30 F-Fodder 0.40 a/ Starch Equivalents. b/ F- stands for crop by-products, C- for processing by-products and K- for oilseed cakes. -67- APPENDIX 2 Table A2.4 ALLOWABLE RANGES FOR THE GRAIN COMPOSITION IN ANIMAL CONSUMPTION (% by weight) Minimum Maximum Wheat 10 15 Corn 10 15 Rye, etc. 5 10 Barley 65 75 -68- APPEI,DIX 3: SENSITIVITY TESTS Technical Parameters 3.01 Although the results of the base solution of TASM-ALV(R) are a satisfactory approximation of actual 1979 conditions, some questions arise as to how the solution for the livestock sub-sector component of the model would respond to: (1) changes in the minima and maxima imposed on each feed group's contribution to the total energy required per head of animal type/technology; (2) different rates of conversion from energy consumed to meat and milk produced; (3) alternative values for the nutritive content of pasture/meadow hay. Six experiments were designed to study the sensitivity of the livestock sub-sector solutions to these parameters. 3.02 Table A3.1 shows the parameter values and selected results obtained. Solutions for the crop sub-sector remained unchanged during the sensitivity tests, with the exception of solutions for the grains which are linked to the livestock sub-sector through the contribution of concentrate (feed grains and processing by-products) to feed group B and through straw to feed group C. The objectiv-e function changed correspondingly as livestock and grain production changed. 3.03 In experiment 1 the efficiency of meat and milk production of 'unimproved' sheep was lowered relative to that of 'improved' sheep. Conversely the efficiency of 'improved' cattle relative to 'unimproved' cattle was lowered. At the same time the nutritive value of pasture was -69- reduced by 14%. The results showed that (compared with the base solution) numbers of 'unimproved' sheep fell and those of 'improved' sheep rose (to the maximum allowed), while in cattle the reverse movements took place i.e., the number of 'unimproved' cattle increased while 'improved' cattle disappeared altogether. Barley production rose by nearly 10%. In experiment 2 the efficiency of 'unimproved' sheep relative to 'improved' sheep was reduced still further and the efficiency of 'unimproved' cattle relative to 'improved' cattle was reduced for milk production. The nutritive value of pasture was restored to its base solution level. As a result of these parameter changes, numbers of 'unimproved' sheep declined further from the experiment 2 levels. There was a sharp fall also in the number of 'unimproved' cattle and strong resurgence in the numbers of 'improved' cattle, compared with the experiment 2 results. Barley production declined slightly below its base solution value. 3.04 In experiment 3 the feed conversion efficiency of 'improved' sheep was raised slightly above the base solution level. The efficiency of both cattle technologies was lowered: proportionately more for the 'improved' technology than for the 'unimproved'. At the same time the maximum contribution to total energy required from feed group B (concentrates) was lowered for cattle while the minimum requirement from Group D (forage) was raised. The nutritive value of pasture remained as in the base solution. The results showed that 'unimproved' sheep numbers declined slightly while 'improved' sheep increased to the maximum number allowed. 'Unimproved' cattle also increased to the maximum allowed but the 'improved' cattle disappeared altogether. There was a slight increase in barley production. Experiment 4 was similar to experiment 3 except that -70- the maximum energy contribution from feed group B for sheep was also lowered and the minimum requirement for 'unimproved' cattle from Group D was raised. The result was that the numbers of 'unimproved' sheep fell from their experiment 3 levels ('improved' sheep remained at the maximum) as did 'unimproved' cattle (improved cattle remained at zero). ,.2 most striking results was that barley production declined substantially by nearly 40%. In experiment 5 the minima and maxima requirements from feed group A were raised considerably for both sheep and cattle and the nutritive value of pasture increased by 50%. This resulted in the same number of sheep as in experiment 3 but slightly fewer cattle. Again the main effect was to decrease barley production, this time to only 36% of its level in the base solution. Finally, in experiment 6 the efficiency of feed conversion into milk and meat was increased substantially for both cattle and sheep. At the same time the minima and maxima requirements of energy from feed group A were reduid below their base solution values while those for group B were increased. No maxima or minima were imposed on energy requirements from group D. The nutritive value of pasture was also set 14% below its base solution value. This resulted in a decrease in the numbers of 'unimproved' sheep below base solution level but an increase (to the maximum) in the number of 'improved' sheep. 'Unimproved' cattle number increased to the maximum, compared with the base solution, but numbers of 'improved' cattle declined to a very low level. Barley production fell by 17% and nearly 25% of the pasture area was unused. 3.05 The results of these experiments can be summarized as follows: (1) The 'improved' sheep technology is mre profitable (in economic terms) than the 'unimproved' sheep technology over quite a wide -71- range of relative and absolute differences in feed conversion efficiency. On the other hand, the distribution of cattle production between the 'unimproved' and 'improved' technologies is quite sensitive to variations in the feed conversion parameters. (2) Consumption of feed group B (concentrates) is sensitive to the relative supply/demand position with respect to group A (pasture/hay). (3) Feed group D (forage crops) is the most expensive (in terms of real resource costs) to supply and it reduces the profitabliity of both sheep and cattle production. Export Prices for Cereals and Livestock 3.06 One question of particular interest with regard to the development of Turkey's livestock sub-sector is the extent to which the production of cereals and livestock for the export market would be in competition for resources. In order to investigate trade-offs which might be involved two experiments were undertaken with parameter settings for the model values representing projection year (1990) conditions with Policy III. In the first experiment international prices (fob and cif) for all cereals were decreased by 40% below the base levels and in the second experiment these prices were increased by 40%. The results are shown in Tables A3.2 and A3.3. 3.07 Table A3.2 shows that with the lower cereal prices the 'improved' technologies in both sheep and cattle production are favored with numbers of animals in these technologies reaching their maximum allowable levels. With high cereal prices, on the other hand, there is almost a complete turn -72- around, with the 'unimproved' technologies, which have a lower minimum requirement for energy from the cereals group, increasing and the 'improved' technologies (except in the case of the intensive cattle technology) decreasing. This switch between technologies is most marked in the case of sheep. 3.08 Table A3.3 shows dramatic differences in the response of different cereals to the movements in the real price of the cereals group as a whole. Wheat and rye production changes substantially as the cereal price is raised and lowered about the base value. Imports and exports of these commodities also change by large amounts. Production levels for corn and barley, however, change very little. Exports of barley under the high cereal prices are 25% higher than under the low prices. Because of the restrictions specified in the model on the proportions of the different cereals in the total animal consumption of cereals, there is little variation in the amount of the different cereals used as livestock feed between the two experiments. 1/ The changes in the real price of grain have very little effect on levels of production and exports of meat and live animals. However, it should be noted that because livestock numbers for certain types/technologies are binding constraints on the solutions to both experiments, the elasticity of livestock production with respect to cereal price (as measured in the context of comparative static analysis) is greater than indicated in Table A3.2. 1/ It is important to note that the model's equations do not say that feed grain has to be consumed by the livestock sector; they only specify * that if grain is consumed by the livestock sector then there are rnstrictions on the mix of the individual cereals. *a Appendix 3 TASM-ALV(R) SENSITIVITY TESTS 011 TECHNICAL PARAMETERS RELATING TO FEED SUPPLY AND FEED CONVERSION EFFICIENCY Table A3.1 Parameter Values a/ Solution Values Pasture Pasture Group A Group B Group D Meat/Energy Milk/Energy Nutritive No. of Sheep No. of Cattle Barley Area a M a M a M Conversioa Conversion Value 'unimproved 'improved 'uniproved' 'improved' Production Used M() (1) (%) (ratio) (ratio) (kg SE/ha) (million) (million) (million) (million) (million ha) (million) Base Solution Sheep b/ 30/20 50/35 5/10 40/50 0/20 80/80 4.4/5.0 11.8/11.1 0.200 40.3 1.0 11.6 1.8 4.2 25.0* Cattle b/ 20/10 35/35 10/15 50/55 0/5 80/80 3.6/4.4 44.4/59.1 Experiment I Sheep: " . - " 4.0/5.0 10.8/11.1 0.172 36.9 2.3' 13.0 0 4.6 25.0* Cattle: - " - " 3.6/4.0 44.4/54.2 Exoeriment 2 Sheep: - " " " " 3.7/5.0 10.0/11.1 0.200 35.0 2.3* 9.5 1.6 4.1 25.0* Cattle: " - " " 3.6/4.0 38.9/54.2 Experiment 3- Sheep: - " " - 4.4/5.3 11.8/11.3 " 38.0 2.3* 13.2' 0 4.3 25.0* 0. Cattle: " 25/35 10/20 - 3.3/4.0 35.0/43.0 Experiment 4 Sheep: " 20/30 5/20 35.0 2.3' 11.7 0 2.6 25.0* Cattle " " - " " " Experiment 5 Sheep: 50/40 80/70 " " " " 0.300 38.0 2.3' 12.3 0 1.5 25.0 Cattle: 40/20 70/50 " " " " Experiment 6 . Sheep: 10/10 30/20 25/30 60/60 0 0 5.5/6.6 14.7/14.1 0.172 38.0 2.3' 13.2* 0.03 3.5 18.8 Cattle: ' 10/10 30/20 25/30 60/60 0 0 4.1/3.7 43.7/40.4 a/ Group A. B. D: m(N) refer to minimum (saximum) proportion of total energy consumed by the animal type/technology which is contributed by the particular groups of feed sources. Meat, milk/energy conversion:. kg of product per kg. of feed (in Starch Equivalents); adjusted by a scaling factor where appropriate for ease of presentation. b/ The figure in front of the slash refers to unimproved' technology, the figure after the slash to improved technology. For example, under the heading Group A - , 30/20 means that the minimum requirement of energy from Group A is 30% for 'unimproved' sheep and 20% for 'improved' sheep. A Denotes that this value is a binding constraint on the solution. -74- Appendix 3 Table A3.2 COMPARISON OF HERD SIZES DUE TO CHANGES IN CEREAL PRICES Cereal Prices in 1990 1979 Levels 1979 levels -40% + 40% Livestock Numbers in 1990 ('000) Sheep - 'traditional' 52,865 57,771 'improved' 3,451* 0 Goats - ordinary 18,598 18,469 Angora 1,835 1,835 Cattle - 'traditional' 19,537 19,848* 'improved' 3,385* 3,095 'intensive' 117* 117* Buffalo 1,560* 1,560* Mule 1,902 3,679 Poultry 88,409* 88,409* * Denotes that this value is a binding constraint on the solution. -75- Appendix 3 Table A3.3 COMPARISON OF PRODUCTION CONSUMPTION AND INTERNATIONAL TRADE DUE TO CHANGES IN CEREAL PRICES ('000 mt) Price Relative to 1979 Pro- Con- Livestock Level duction sumption Feed Exports Imports Wheat -40 7,593 11,207 1,059 0 4,673 +40 12,013 9,473 1,046 1,494 0 Corn -40 2,235 647 1,588 0 0 +40 2,140 571 1,569 0 0 Rye -40 403 400 529 0 526 +40 2,230 400 523 1,307 0 Barley -40 10,550 1,140 7,412 1,998 0 +40 10,846 1,024 7,322 2,500* 0 Sheep & Goat -40 372 321 - 51 0 meat/stock +40 364 319 - 45 0 Cattle & Buffalo -40 369 262 - 107 0 meat/stock +40 366 261 - 105 0 * Denotes that this value is the upper limit on the first 'step' of the export demand function. Beyond 2,500,000 mt, barley exports are sold at a price 1/3 lower. -76- Appendix 4: MISCELLANEOUS DATA 4.01 Table A4.1 compares land use solutions of TASM and TASM-ALV(R). Irrigated area is binding for all solutions, including in the projections for 1990 where the maximum area available has been increased by 27%. Rainfed land, however, is not binding in any solution. Although the rainfed area increases 31% and 26% respectively for TASM and TASM-ALV(R) under Policy III in 1979, the projections for 1990 indicate that this area actually decreases by 17% for both versions of the model. The decrease in the use of rainfed land is mainly due to the improvement in cropping techniques which require higher inputs and produce higher yields. Pasture area is binding for TASM-ALV(R) solutions and non-binding,for TASM's. 4.02 Table A4.2 shows the limits on quantities exported for different commodities under Policy III. Note that wheat and barley are allowed to export another million mt each but their prices beyong the 2.5 million mt are reduced by 33%. 4.03 Table A4.3 shows the domestic resource costs (DRC) obtained from TASM-ALV(R) base solution at TL47. DRC is the ratio of non-traded inputs over value added at border prices. DRC's for TASM-ALV(R) in general are much higher due to a higher foreign exchange (therefore lower value added at border price). It is interesting to note that sugar beet, wild cherry and soybean are not competitive internationally under TASM-ALV(R) base solutions. APPENDIX 4 Table A4.1 COMPARISON OF TASH AND TASM-ALV(R) LAND USE SOLUTIONS BASE YEAR (1969) PROJECTION YEAR (1990) Maximum Available Base Solution Policy III Solution Maximum Available Policy III Solution TASM TASM-ALV(R) TASM TASM-ALV(R) TASM TASM-ALV(R) TASM TASM-ALV(R) TASM TASM-ALV(R) I AREA ('000ha)a/ - Rainfed 22,607 19,353 16,218 14,993 21,220 18,947 21,857 18,603 13,406 12,451 . Irrigated 2,794 2,794 2,794* 2,794* 2,794* 2,794* 3,544 3,544 3,544* 3,544* Tree Crop 2,749 2,749 2,279 2,279 2,369 2,369 2,749 2,749 2,183 2,178 Pasture 21,746 25,000 20,377 25,000* 20,359 25,000* 21,746 25,000 20,409 25,000* Total 49,896 49,896 41,668 45,066 46,742 49,110 49,896 49,896 39,542 43,173 AREA (% change from base solution -- 1979) Rainfed +31 +26 -17 -17 Irrigated 0 0 +27 +27 Tree crop +4 +4 -4 -4 Pasture 0 0 0 0 Total +12 +9 -5 -4 a/ Net cultivated area (except in the case of pasture), i.e. multiple cropped land is counted only once. * Level of use is at the maximum possible, i.e. the availability of this type of land is a constraint on the solution variables. -78- Appendix 4 Table A4.2 LIMITS ON QUANTITIES EXPORTED UNDER POLICY III ('000 mt) Wheat a/ 2,500 Barley7a/ 2,500 Chick pea 250 Dry bean 250 Lentil 250 Potato 250 Onion 250 Green pepper 250 Tomato 250 Cucumber 250 Cotton 200 Tobacco 200 Citrus 500 Grape 50 Apple 200 Peach 200 Melon 50 Strawberry 50 Quince 50 Pistachio 50 Hazelnut 50 Tomato paste 50 Raisin 100 Shelled hazelnut 200 a/ Wheat and barley are 'allowed' by the constraints specified to export up to another million mt each over and above the 2,500,000 mt limit, but at two-thirds of the price. 79- Appendix 4 Table A4,3 DOMESTIC RESOURCE COST RATIOS (DRC'S) FROM TASM-ALV(R)-IN 1979 TASM-ALV(R) TL47/US$ Wheat .528 Corn .904 Rye, etc. .523 Rice .533 Barley .361 Chick pea .576 Dry bean .534 Lentil .421 Potato .317 Onion .767 Green pepper .364 Tomato .599 Cucumber .302 Sunflower .537 Olive .538 Groundnut .667 Cotton .807 Sugar beet 1.191 Tobacco .756 Tea .058 Citrus .666 Grape .675 Apple .665 Peach .669 Apricot .466 Cherry .727 Wild cherry 1.016 Melon .518 Strawberry .673 Banana .060 Quince .729 Pistachio .276 Hazelnut .760 Soybean 2.317 Sesame .203 -80- REFERENCES Gencaga, H., "Animal Husbandry Strategies for Turkey," paper presented to the 7th National Livestock Conference, 1981. Le-Si, V., P.L. Scandizzo, and H. Kaonakoglu, "Turkey Agricultural Sector Model," The World Bank, March, 1983. McDonald, P., R.A. Edwards, and J.F.D. Greenhalgh, Animal Nutrition, Oliver & Boyd, 1973, 2nd Edition. State Institute of Statistics, Agricultural Structure an,d Production, 1980, Ankara, Turkey. World Bank, Erzurum Rural Development Project -- Project File C12. , Fourth Livestock Development Project, Supervision Mission, October-November, 1981: "Revised Farm Development Models." , Turkey - Agricultural Development Alternatives for Growth with Exports, January, 1983, Report No. 4204-TU.