Raising the productivity of small farmers - Tanzania case study, 1975

This internal working pwper is prepared for STAFF USE ONLY. The views expressed are not necessarily those of the World Bank. RAISING THE PRODUCTIVITY OF SMALL FARMERS TANZANIA CASE STUDY, 1975 AGREP Division Working Paper No. 33 Prepared by: K. Friedrich (Consultant) Economics and Policy Division Agriculture and Rural Development Department March 1980 TABLEF OF COFrTETS Pae I. INT'RDUCTION 1.1 Background 1.2 The Small Farmer in Tanzania 2 1.3 Objective: of the S-udy 3 1.4 Mýethodology and Approach )oc 2. DT TlSSAQFIIG 2.1 Farm Economics in Various Study areas 2.1.1 Coffee-Banana Farmin.g in Kilimanjaro 2.1.2 Nlaize-Beans-Ve&etable Farming in Lushoto 4 2.1.3 Kaize, Coffee and Pyrethr"um Farming in Mbeya 2.1.4 Cashewnut, Coconut and Rice Farminz, in the Coastal Atea 51 2.1.5 Maize and Rice .Farming in MoroÅroro 2.1.6 !,.aize and Co ton ar,in in Shinyinnrr. 2.1.7 Maize, Groundnuts and Tobacco 'Farmning in Ta,,cra 10: 2.1.8 Maize, Pulse and Groundnut Farming in Iringa 122 2.1.9 Cereal and Tobacco Farming in Ruvuma 13 3 2.1.10 Cereal Farming in Singida APPEIDIX A SUREY AEAS UITH P2RTIAL I23Ii.TOF:LTICN - 164 APPE'IX B SUPPLB'D7TARY TA3LES SOITG CROP LABGUR REZUIP NTS BY RZGION 68 I. INTRODUCTION It is now wid.ely recognised-that ii t .e current b2 -be -effort to rapidly increase food production to feid the world' s popula .on, sall-.fax -rs- in -the less developed countries hold the key to success. Basic lly, this is Y ,cause they constitute the overwhelming majority of the world's food I -oducers. Bu' their per cloita output is disapointingly low, and i!ays and means mu. be found to 'aelp them enb .nce their productivityE What has become quite clear over the las+ several yea s is that there is a fundaiental lack of knowledge about the small farme' - about hi objectives, resource base, ocnstraints, priorities, etc., and t-is has criti-Llly hampere, effective design and implenontation of programmes aimed at increasing hi output and -.ncome. In recognition of this need to obtai), a clearer a,,;sessment of tae world situation with regard to the small farmers, and to develop a fra;iework for fo mulating effective pro,grammes geared towards br%nging the smill farmer population mos' effectively into the development process, FAO and IBRD agreed to jointly undertake his research project entitled "Raising the Productivity of tia Small Farme s". The project has several componer's. Part I inolved a summay of the 1970 agricultural census data. This par provides infoimation on the Issic economic characteristics of farm and trends ii the timber of small farmers since 1960. Thus, the magnitude of the small farmer proolem on t world basis is determined in quantitative terms. Part II builds on information obtained tn Part I. The main objectives of the second part were to: - provide a muantitative analysis of th small farmers' economic position, their efficiency of resource use, thei-c income position, and the effect of factors external to the farm on their 1ncome situation; - estimate the potential growth for the viall farmer sector %ith regard to improved production and income; - identify constraints to increased produetion and income and indicate ways to remroving them, thus permitting the eoi'ated poten''al to be achieved; - propose feasible investment - oriented development programmes geared towards improving the relative position of small farmers. The Gcver.;errnt of -re C: d Republic of =anzua was approached and it ~r~e to havt Prt 'CZ o the uuroject a oxt i t1 ouin-ry Ž 2 The Riall Farmer i.n Tanz.'i a Csto rul- "f i e .witu t.1 ccmu.l of Tanz=ia of 1971-72, in the bsisteC e soo"i, 2(! x:illt farmers id 2,9' -nillion hectares of far:= Iand or .1 h o.res on aVCrage "4äe fall-:ing d:.stribuu .n of fQrzmers bN7 a.ferent -aE: size classes indicates ~he s ¿ divrgen-:cies; alIost 60 percen-: hold farm sies Jess than one hectare, Rogl_>h) F4a=mers 'poecen+,) cr 0-50 31.5 50 1,00 26.7 O0 -2.0 24.7 l00 - 3 00 8,9 3 0 - o0 0.4 00- 5,2 02 - 00 ané over 7,6 T2e-al 100.0 With the exc;f':ý of th:e 7,6 perce.nt more priviledged farers, all others could ScId ':i' '-'ith regard to their f1rm size . A further delinea:uion according -a -r .able seems ±naoDropriate alt iough i-', would have the inherent appeal of nrm.ece a sebecaise ofr2 d dpj'u aviaii l t -.ould, how;ever, benarthe da-n.Eor o vrsmpliying the problem. .3pec:al'y as it fails to meaaure the econoric s o fira uszinesses. The determi n -io - )f the latter is important because it r'rt-sens j- cpirnd 'easureenx of b-,.h the resource base- including far- land, and ThXe licienc WiM,h whch tt is used. rhe ecouroic size or smalln3ss o. farms could thus .e established using the returs frafa :g s deS erares. This *nforration is, however, only available after careful anasis of an one far- 'busii ass anc. Emzall fk-rmers can only be identified fror:. -n a, -:he end ofc* a. ie i. -.3- Smallness of a farm can also be measured by other criteria, such as total value of farm assets or nutritional satisifaction, but the cost ideal measurement seems to relate to the amount of cash left over after completion of the production period, which-can be used for investment, replacement or savings. This ma6nitude should be viewed in conjunction with the farmerls change in net worth during the same period. This concept coincides with the concent of 'viability' of farms, which has becn defined, e.g. by Lindsey as the 1/ ability of a fari. to re-organize - . It implies that the flow of income exceeds that recuired to cover living expenses, current operating expenses, depreciation and annual debt obligations. Small farms, reogardless of size of resource ownersaip or resource availability, are set equal with. 'not viable' farms, lacking the re-orcanizational capacity, assumingacertain basic standard of nutrition and" living. The task is to adoptionthem and to make them viable. l13 ObJ1ectives of the Study As mentioned in section 1.1, a realistic and effective `development strate&r can be formulated only if there is a f`undamental understanding of existing conditions, including interpretation of production relationships and behaviour at the producer level. This ensures the most efficient resource allocation and performance. The basic objective of this study was to gain this detailed knowledge through collection and analysis of micro-economic, farm mangement type information from farm cazples selected from different a,ro-ecological zones of Tanzania. The underlying hypotheses were the following: 1. Farmers, as a group, make optimal use of resources available to them, given the constraints under which they operate. 2. Some of the farmers will have evolved technicues or approaches which minimise the effects of major constraints or oroduction and, therefore, achicve ireatLr production and income per farm unit. 3 These techniques or approacnes evolved by some of the farmers are directly transferable to the majority of farmers in similar agro-ecological and socior economic settings. However, it is also possible that ad; itional improvements in small farr;er production and income can be achieved through adoption of new (selective) technoloey. 1 Source: Bureau of Statistics, Dar-es-Salaam. Thus, detailed information -.as requi . on all aspects of the small farmer's production process, including resource ava;1ability and utilization, income and consuwption, constraints to increased prod-ction and income, technology of production, etc., to facilitate the testing of the abcve hypotheses and formulation of realistic development proposals. 1.4 Yethodol& eand Approach Althou7!h a number of farm management type studies have been carried out in the pas- in Tanoania, for none of them is the primary farm data available,neither is the information systematically coordinated to be used in its aggregate form in which it is available. Also new developments may have occurred since the introduction of new enternrises, change in production structures, etc., which may have changed the original farm organioation in some areas. A new and more comprehensive, as well as a mlore systematic data base seemed necessary to meet the abjectives of the present research st Lidy The 1970 census of agriculture offered the most suitable frame for the selection of farmers to be investigated. However, bearing in mind the need for inter-relating farm economic information with agro-ecological data and zone mapping, 20 major small farmer/farming type areas (study areas), each contained in a totally specific agro- ecological zone, were defined as a first step, A total coverage of the whole country was rejected as impracticable because of the great dispersion of survey farms and also because of the substantial logistic difficulties. The decision as to the number of study areas, as well as other quantitative detail, was reached in accordance with resource availability, mainly in terms of enumerators, their capacity and time. The choice of location of study areas was reached by discussions with those people familiar with the country so as to ensure that all the major agro-ecological zones were covered. As a second step, the 20 cle y area (see Mao 1) were converted to Enumeration Areas (EA's) covered by the census of agriculture, Each study area comprised between 10 and 15 EA's, out of which 90 to 100 sample farmers were selected at random. For the data collection, 20 students from the Faculty of Agriculture, University of Dar-es-Salaam, were employed, each of them being allotted one survey area coinciding with his home area, with which he was familiar. -5- The student enumerators were briefed thoroughly in the use of the survey forms, as well as to the best approach to gain the farmer's cooperation and confidence, a pre-requisite for obtaining reliable results. Effectively -j months were availabe for the field work and the only means of transport possible was publico Two supervisors (tutorial assistants of the Faculty) were originally envisaged as being attached to the study. Due to unforseen problems this proposal did not materialise in time. However, due to the anormous logistic problems it is doubt .-I whether their employment would have substantially affected the outcome of the investigation, Contact and supervision during the field work was coordinated by the Department of Rural Economy, Faculty of Agriculture. As it turned out, two areas (112, 119) were not surveyed at all; a third area (104) was discarded from the analysis; from seven areas only partial information became available (103, 106, 107, 108, 110, 112, 115); while from ten areas complete sets of data were obtained and were available for analysis. These were: 101 Kilimanjaro 111 Morogoro 117 Iringa 102 Lushoto 114 Shinyanga 118 Ruvuma 105 Mbeya 116 Tabora 120 Singida 109 Coastal Area The survey-forms used were variants of FAO's Farm Management Data Collection and Analysis System (FNDCAS). These forms were streamlined as much as possible to suit the general conditions of the country. Due to lack of time during the preparation phase, it was not possible to develop specific sets for each survey area and its particular agriculture. Reliance therefore was heavily placed on the capability of the student enumerators, especially with regard to the constructing of complete input/ output data for each individual enterprise. The survey forms (with the exception of questions relating to the farmer's opinion and attitude, and the coverage of Ujamaa Villages) were constructed in such a way that after scrutiny is applied and coding compLetei, the data can be transferred directly onto the computer input means (cards). -6- MAP 1: SURVEY AREAS O, THE FAO/IERD S- IL FRME STUDY, TANZAnIA Administrative Areas 1967 k) 103 13 112 0114 ý"104 120 10 S116 119 111 117 05 118 110 101 Kiliinanjaro 108 Tanga 115 Mara 102 Lushoto 109 Coast 116 Tabora 103 West Lake 110 twara 117 Iringa 104 Wigoma 111 Morogoro i Ruvuma 105 Mbeya 112 Sukua 1 Dodoma 106 Arusha 113 Mwanza 1-0 Singida 107 Kondoa 114 Shinyanga 7- The technical coverage of the survey )rms by which both individual enterprises and the farm were examined, was as folloirs: 1. Family composition and permanent abour 2. Availability of farm land and its charateristics. 3. Farm land utilization (annaal and perennial crops). 4. Provision to cover'machinery and Iraft animals, if and when they occur. 5. Farm buildings and small tools. 6, Inventoxr of productive livestock. 7. Financial liabilities/indebtedness. 8. Inventory changes. 9. Input/output data for each enterprise (crops and livestock). 10. Household consumption and expenditure. 11. Farmer's opinion and attitude, felt constraints. 12, Ujamaa (as part of the farmer's economic activities). With regard to analysis, data were processed and tabulated by above mentioned computerized systems. Thus, e.g. basic farm statistics, gross margins for all entdrprises and for the farms were calculated. Subsequently, both enterprise and farm data were stratified according to major parameters, such as, farm size, farm returns, fertilizer application, mechanization, etc., with the aim of identifying certain sub- strata of the population which are different from the total and to analyse the causes and results of these differences. Production functions were applied. Multiple step- wise regression of the Cobb--ouglas type, which is linear in the logarithmic form, seemed the bett method to analyse the aggregate farm data as well as specific crop and livestock enterprises. The definition of marginal returns for all production factors entered in the regression allows for an approximation of their optimal combination for maximizing faring results. Farm programming was attempted through linear programming, which under the restraints imposed, indicates equally optimum farm organizations. -8- 2. DATA ATATtIS AND F LIGZ 2.1 Farm Economics in Various Study AreP3 2.1.1 Coffee-Banana Farming in Kilimanjaro The Kilimanjaro survey area is delineated by the divisonal censu areas 061i 0612 and 0613. 74 farmers were surveyed and the results are prosented 'elo7. i6 earlier survey of coffee-banana-farming was carried out by BECK In Aj: ai ±s summarized by RUTHERNBERG along with other relevant data. The Kilimanjaro farmer is 47 years of age and is head of F,6.9 he.. 9miFly and 0.2 permanent hired labourers. He farms 1.18 hectares, 2/3 in coffeea.ba an'. .1/3 in maize-beans and other crops. Better farmers have farm sizes of alnost t:ine The average size but with similar cropping pattern. Only the portion of coffee-banana, is slihtly increared. Livestock numb3rs kept average 1.5 head of cattle. est farmers keep 2,,2 head. Values of production amount to 7,715 shs per farm on average 7ad are more thau doubled in best farmer. S* Values of =roduction are made up to about half by coffee-banana, 33 percent by livestock and the remainder '6r maize- beani and other crops. Best farmers have both higher contributions to the val'ue of production from coffee-banana and livestock, while variable costs are not proportionately higher. Variable costs which average 586 shs per farm are spread over . witde spectrum of items, which do not show any definite trend between farm groups. Labour npta amount to 2,093 hours per farm, spread on average to 40 percent for crop enterprisQs and 60 percent for livestock. Since the structure of production and costs does not inag proportionately between farm groups, the same relative differences ex:ist -Tith regard tc farm income which averages 7,129 shs ranging between 2,076 shs and 18,203 sh per farmo Only the 20 percent be8t farmers achieve farm incomes significantly above th aluated average levels. Off-farm income amounts to 952 shs and. is higher in tr-ese farm groups with lowest farm income, thus narrowing the gap of family incomes between farm g-rops. 1/ Beck, R.S. An Economic study of Coffee-Banana Farms in the Machame Central areas, 1961, mimeo. 2/ Ruthenberg, H. Coffee-Banana Farms at Mt. Kilimanjaro, in RuthenberS, H Sma-.'3L Holder Farming and Small Holder development in Tanzania, MInchen, 'c68 - io - TA.BLE 2.1: Composite Characte.,tkics of KilimanoL, _O_F M.21 -rn goyp a' or :,2A -1 Levol cf farm rrodtiction Details Average Very Very high HeCiml Low INo. of observations 15 tFarmerv family, labour Yamersl age, years 40.7 47 - 1 9 43.9 1, 47 3 47-1 Farmers' education, school yrs 2.9 3 , 5 2:2 4:5 2:5 3-1 !Farm family, people 605 7 0 6.4 7 1 7 1 6.9 I Perm. hired labour, nos 0.2 0:1 il 0-1 0-4 1' 0:4 0.2 (Total people on farm 6-7 7.1 6-5 7.5 7-5 7-1 [Labour availability., man- .4 e SLaival ent s A.6 1 3.6 3.8 4.0 FAF16 LAND Farm size, ha 2.o6 1.28 1-15 0-75 0 70! 1.18 (standard error) (0-34,k) (1-17)1 (0.16)1 (0.09) (0:00 (0.10) Cropping_pattern, % Coffee-bananas 70.0 76.o 64-0 50.0 1 66.0 Maize-beans 30.0 24-0 if 32 2 34-7 35 7 30-5 Others I la:9 1.3 14:3 -305. TOTAL 100.0 100.0 hoo.0 100.0 100.0 100.0 Livestockf head of cattle 2.21 0.93 2-47 0 6o 1.46 1-53 Variable inputs, shs 805 723 638 i59 419 586 Percentage distribution: Seed 14.9 21.0 15.7 14.8 19.8 17-0 Fertilizer 18.3 21.3 33-1 35-9 28.2 25.8 Pesticides 16.7 22.1 26.6 29-4 26-7 23-1 Hired labour 23.3 21-3 12-4 7-7 9.1 16.6 Hired power - - - - 5.6 0.7 ,Others 12.9 13-4 8,6 12.0 lo.6 11.6 Livestock expenses 0.9 3.6 0.2 - 5.2 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 Labour-inputst hrs 31021 2,301 21576 11383 11248 2,093 Crop production 34.3 37.1 39.9 52.2 47.6 4Oe3 Livestock production 65.7 62.9 6o.1 4 .8 52.4 59.7 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 Value of farm production, shs 19,oo8 8,04 59716 4,005 2,485 7,715 Tstandard error) ",'6 59 (2 57 80) (10 -L .7 4) (189.82, (11088.98) Percentage distribution: Coffee-Banana 56.3 57-4 55-1 59-7 56.o 52.2 Maize-beans 5.8 9-7 10.2 17.6 24.3 9.8 Others - - 4.9 2.0 - 4.6 Livestock 37-2 32.9 29.8 20-7 19-7 1 33.4 TOTAL 1 100.01 100.0 1 100.0 100.0 100.0 100.0 - 11 - TABLE 2.1: (continued) Detail Level of farm production Average Very Very -high High Medium Low low Value of farm production, shs 19,008 8,084 5,716 4,005 2,485 7,715 Variable costs, shs 805 723 638 359 419 586 Sum of aross margin, shs far come) 18,203 7,361 5,078 3,646 2,076 7,129 10ff-farm income, shs 670 874 540 1,108 1,608 952 lFamily income, shs 18,873 8,325 5,618 4,754 3,684 8,081 Home consumption, shs 10,178 4,260 3,206 2,375 1,376 4,199 Household expenditures 3,765 3,650 2,840 3,017 3,013 3,747 Cash for investment, replacement SavinGs, consumption 4,930 325 -428 -638 -605 135 LANTD PRODUCTIVITY Value of production, shs/ha 8,836 5,750 4,415 4,861 2,965 6,041 Variable costs, shs/ha 390 564 554 478 598 496 Gross margin, shs/ha 8,446 5,186 3,861 4,383 2,367 3,545 Labour input, hrs/ha 503 667 894 962 848 715 LABOUR PRODUCTIVITY Value added, shs/man equiv. 3,997 1,833 1,433 966 528 1,762 Value added, shs/labour hr used 6.08 3.26 2.00 2.65 1.69 3.45 Ratios Value of output to value of input a. Labour not costed 23.6 11.2 8.9 11.2 5.9 13.2 b. Labour costed 4.9 2.7 1.8 2.3 1.5 2.9 Degree of family labour use, % 41 35 46 25 21 32 Degree of commercialisation, % 47 48 44 41 45 46 Farm income as percent of total income o96 89 90 77 56 88 Cash balance as percent of total income 26 4 negative 2 - 12 - The Kilimanjaro Farm The Kilimanjaro coffee-banana farmer is on average 47 years of age and has an educational standard of.only 3 school years. 5.8 family members and 0.2 permanent hired labourers plus the farmer constitute the household of 7.0 people with a potential work force for agriculture of 2.8 man-equivalents. 17 percent of all farmers are engaged also in off-farm employment and earn for a 213 day input 15 shs per day. This would amount to 520 shs/year on average for all farmers. The Kilimanjaro coffee-banana farm is on average 1.18 ha. In only very few instances is farm land rented (shs 150/ha/year) in addition to land held in customary ownership-like possession. No ujamaa farming fell within the sample of the survey. Equally no fallow land is recorded, thus setting the size of cultivated land equal to the size of farm land managed. Farm land is distributed as follows: Farm size group ha Average farn Percent of Percent of size farms total area Under ,50 0.28 16.2 3.9 *50 - 0.99 0.76 35.2 22.7 1.00 - 1.49 1.18 21.6 21.7 1.50 - 1.99 1.68 10.8 15.4 2.00 - 2.49 2.19 10.8 20.0 2.50 and over 3.56 5.4 16.3 1.18 - 100.0 100.0 The above structure suggests a rather equal distribution, where only the largest farms have distinctly more land at their disposal than the rest. The cropping pattern is the simplest of all survey areas: coffee-banana is grown on 0.77 ha and maize-beans on 0.36 ha. Other crops grown in purestand or other combinations account for the rest 0.05 ha. - 13 - Coffee-Banana Mixtures Almost all farmers grow this crop mix, whose basic parameters are summarized in Table 2.2. Coffee yields per hectare amount to 307 kg/ha or 0.24 kg/tree, valued at 5.10 shs/kg. Coffee contributes slightly more than 40 percent to the value of production. Per hectare values of production amount to Shs*4,374. _1/ Production is more than double in the best performing farms, but it is achieved on very small cultivation sizes with higher coffee and banana densities and also higher levels-of all other inputs. As productivity (value of production/ha) decreases, so does the use of all inputs. TABLE 2.2: Input/Output Coefficients Grouped by Level of Productivity - Coffee-banana, Kilimanjaro - Input/output coefficients Level of produtivity Average Very Very high High Medium I Low low No. of observations 14 14 141 14 14 70 Cultivation size, ha 0.29 0.29 0.64 1.04 119 0.82 Coffee trees, nos/ha 1,692 1,774 1,914 1,010 617 1,269 Banana clumps, nos/ha 1,535 1,607 7241 665 445 1,001 Variable inputs, shs/ha Pertilizer 186 87 106 86 70 92 Pesticides, shs/ha 227 98 174 108 78 110 Others 333 160 196 155 134 175 TOTAL 746 345 6 349 282 377 Labour inputs, hrs/ha Wedding 351 125 196 161 109 258 Others 1,727 536 662 680 282 571 TOTAL 2,078 661 858 841 491 829 Value of production, shs/ha Variable costs, shs/ha 74 345 476 349 282 377 Gross margin, shs/ha 8,759 5,401 4,045 3,260 2,369 3,997 1 This includes an estimate of home-consumed banana production assuming half of the clumps producing one bunch a year, valued at 5 shs each. - 1/1 - A Cobb-Douglas production function using the value of production as the dependent variable resulted in 3 positive elasticities only. These were for land 0,52* 1 coffee trees 0.18 and banana clumps 0,35*, Other variables (labour in weed control, fertilizer and pesticides) had negative and insignificant elasticities. The sum of elasticities was 0.98 and the adjusted R2 0.65. The marginal product for land was found to be 2,274 shs/ha, compared to its -stimated cost of 1,968 shs/ha 2/, the ratio of the former to the latter being slightly larger than unity. Marginal products for both coffee and bananas are positive, although at such lot- levels as 0.61 and 1.52 shs/ tree respectively. These figures have to be compared with current establishrlent costs in order to determine whether expansion would be profitable. In case additional establishment takes place, it would seem advantageous to increase the proportion of bananas, rather than that of coffee. For coffee-bananas the best way to increased returns would be by slight area expansion, although results would not be spectacular. Maize-Beans Mixtures Productivity of maize-beans, which is grown by 75 percent of all farms, is only half that of coffee-banana, although the gap is small for best performing farms. Compared to coffee- ananas, mach higher levels of all matPrial inputs, as well as labour inputs for land preparation, planting and weed control are used (Table 2.3). As productivity decreases all inputs decrease, with the exception of the cultivation size, which continuously increases. The result of the Cobb+Douglas production function analysis gave for labour in land preparation and planting an elasticity of 0.46* implying relatively marginal returns. Other variables (with theexception of weeding labour and pesticides which show negative elasticities) also yielde- significant coeffici nts: land 0.22, fertilizer 0.19, power input 0.10, and seed 0.26. The sum of elasticities amounted to 0.96 and the adjusted R2 was 0.90. 1 * significantly different from zero at the five percent level. 2/ 10 percent of farmer's estimated land values. - 15 - TABLE 2.3: Inout/Output Coefficie its by Lvel of Prodcotivil% - Maize-I3ans, Kilimanjaro - Lev(l of productiv ± Input/Uutput coefficients v rcS Very.y . high High Medirm Tn- . No. of observations 11 11 11 11 1 Cultivation size, ha 0.12 0.34 0.59 0.58 0 77 Variable inputs, shs/ha Fertilizer, shs/ha 528 229 105 113 3 Pesticides, shs/ha 305 123 39 93 0 Seed, shs/ha 799 231 126 178 01.1 Power, shs/ha 1 ;8 - Others, shs/ha 238 84 78 62 93 7C TOTAL 1,918 667 - 348 446 315 Labour inputs, hrs/ha Land prep, planting 847 401 228 240 201 27 T'--d control 338 154 75 94 94 13 Others 524 218 160 169 106 1/ -TOTAL 1,709 773 463 503 40 53 Value of proziuction, shs/ha Variable costs, shs/ha 1,918 667 348 446 315 j Gross margin, shs/ha 5,703 2,416 2,122 1,085 580 1 58 53 1/ Tractor ploughing is employed by four percent of the farmers only. Average products were high and, with the exception of farm land, marginal returns were substantially higher than the opportunity costs of all inputs having positive elasticities. No significant contributions to returns are expected from additio-aal application of pesticides and labour in weed control. Marginal-returns-to-opportunity.- cost-ratios indicate that cultivation should be more intensified, that is, more labour should be devoted to land preparation and planting on relatively smaller acrea6es, more seed and fertilizer could profitably be used and subFtantially more I,-n prart averagc power inputs should be used. For the mixture of maize-beans these findings can be substantiated. An Ed uptment of the proportions of the different inputs to the points where all ratios are as close to unity as possible, at same present cost levels, results in a significantly hghcr - 16 - .f pcroduction. This is achieved by a much more labour and capital intensive mode iT ltivaticn. The optimum resource combination is given with 0.11 ha of land, "i labour hours in land preparation, 183 shs for fertilizer, 91 shs for Dover use at 250 shs for seed. At srme as present total costs of 1,248 shs, the value of p:ouctOion amounts to 4,020 shs. Although these input/output levels are only ix:1icative, this mixture of maize-beans can `e grown more profitably than at present, _.nticularly if care is taken to ensure that an efficient combination of inputs is c'hiev,Dnd. 50 percent of the surveyed farmers keep cattle, on average of 1.7 head, at shs 504/head. The labour input amounts to 1,262 hours on average per farm, being mainly spent on herding. Variable inputs of shs 35 are spent for salt and inerals. The value of production from cattle keeping amounts to shs 3,195, forthcoming >alf and half in cash and kind. Inventory changes are not included in the analysis. Cobl-Dorglas production function analysis produced an extremely low adjusted R2 ::arginal returns are not high enough for both the head of cattle kept and the laiour amployed. Variable inputs (feed, minerals, etc.) are employed too rarely, as compared to their opportunity costs. In addition to cattle, 11 farmers keep sheep, 2 keep goats and 16 keep poultry. For these types of livestock, inputs, including those for labour could not be estimated. Output amounts to shs. 472 for sheen, shs.338 for goats and,shs-729 for poultry keeping farms. The contribution of these livestock enterprises, on average, over all survey farms amounted to shs.237 or 3 percent of the total value of production of the farm. WThole farm For the coffee-banann farm of Kilimanjaro area the average value of production amounts to 7,715 shs. Two thirds is derived from crop enterprise and one third from livestock enterprises. 1/ In another distribution, 45 percent is cash and 55 percent in kind (home consumption). Total variable costs amount to 586 shs/farm, leaving a sum of gross margin of 7,129 shs/farm, which is more than in any other survey area. Labour innuts amount to 2,053 hrs per farm, spent at about 0 percent in crop production (see Table 2.4). 7 'TNot including inventory changes. - 17 - The 20 percent best performing farms edrn 2- times the average returns, with higher absolute levels of all in uts, including farm land. They equally achieved the highest level of productivity (returns/ha). When grouped according to farm sizes, values of production/farm as well as gross margins increased only slowly 1rith increasing farm size, but rise rapidly in the highest farm size groups. This trend is, however, not followed by other variables, notably labour inputs, which seem to rise in a more linear manner. Cobb-Douglas production function anal:rsis produced a similarly high elasticity for land (057*) as for the coffee-banana mixture. In addition, labour for livestock showed a high elasticity (0.28),- but all other variables included in the analysis produced positive but small elasticities. These gere labour for crops 0.06, number of livestock 0.02, variable costs 0.02. The sum of elasticities amounted to 0.95 and the adjusted R2 was 0.85. With the above elasticities the magnitude of marginal products are already pre-determined, which for land, are in excess of likely opportunity costs. For other variables the comparison of marginal returns to opportunity costs does not produce any meaningful picture. This is because, as has been shown in the foregoing, the two major enterprises (crop mixes) show clearly different economic patterns: Coffee-banana Maize-bepns - low returns to all labour - high returns to labour (land preparation) - high returns to land - low returns to land - low returns to all variable inputs - high returns to fertilizer, seed, power - enterprise should be expanded with labour - labour and capital should be more and capital used less intensively. intensively applied. As a result of this, farm analysis by Cobb-Douglas function produced an overlay of oth enter-rises an-` this tcnded. to ,ist-_rte' results. Nevertheless, the cverall indication seems to be that farms shoulI be exTcn,e' to permit a more cntimal usage of other resources. The average farm income (sums of gross margins) amounts to shs.7,129 per farm. Grouped by level of farm production, farm incomes show a wide variation between farms; and range from shs.2,076 in the lowest to shs.18,203 in the highest group (Table 2.5). This is to some extent compensated by farms with smaller farm incomes having a larger proportion of off-farm income (labouring 'r contribution from other Family members, pension, etc.). Thus the family income amounts to shs.8,081 on average, ranging from - 18 - TABLE 2.4: (a) Input/Output Coefficients for Farm by Level of Productivity - Whole farm, Kilimanjaro - Level of farm production Input/output coefficients VeryAverage __________Very_________ ____MeimVery Avre high High Medium Low low No. of observations - 5 15 15 - 15 IA 74 . Cultivation size, ha Coffee-banana, ha 1.61 1.03 0.74 .0-51 0.45 0.87 Others 0.45 0.1 _ 0.41 0.24 0.25 0.36 TOTAL 2.06 1.28 1.15 0.75 0.70 1.18 Cattle, head 2.2 0.9 2.5 - 0.6 1.5 1.5 Labour inputs, hrs Crop prod., hrs 1,115 53 1,028 721 528 831 Livestock, hrs 1,906 1,548 1,548 662 720 1,262 TOTAL 3,021 2.301 2,576 1,383 1 24,8 2,093 Value of production, shs 19,008 8,084 5,716 4,005 2,485 7,715 Variable inputs, shs 805 723 638 359 419 586 Gross margin, shs 18,20- 7,361 5,078 3,646 2,076 7,129 (b) Input/Output Coefficients for Parms Grouped by Farm Size - Whole farm, Kilimanjaro - __Fcrm size -roup, ha Input/output coefficients Under .50 - 1.00 - 1.50 - 2.00 - 2.50 and 0.50 .99 1.49 1.99 2.49 above. No. of observations 12 26 16 8 8 4 Cultivation size, ha 0.28 0.76 1.18 1.68 2.18 3.56 Cattle, head 1.2 1.5 0.75 1.5 3.5 2.0 Labour inputs, hrs Crops, hrs 553 759 795 1,051 1,198 1,339 Livestock, hrs 1,173 1,145 974 1,252 1,375 2,993 TOTAL 1,726 1,904 1,769 2,303 2,573 4,332 Value of production, shs 4,330 5,501 5,489 8,639 10,480 33,666 Variable costs, shs 265 512 618 867 766 980 Gross margin, shs 4,075 4,989 Z,871 7,772 9,714 32,686 - 19 - shs.3,684 to 18,873 per farm. From these values the imputed value of home consumption is deducted as well as household exoenditures in order to derive the family's cash balance. The latter can be used for investment, replacement of capital goods, savings or increased consumption and is an indicator of the viability of farms. Home consumed farm products were valued at ahs.4,200 on average, but even home consumed goods vary with the level of farm production. Consumption in higher groups is much above the average of the lower groups, whereas purchased consumption items are rather constant over all farm groups, and averages 3,7A7 shs per farm. On average a positive, although small, cash balance of shs.135 accrues. Comparing farm groups, however, it becomes obvious that only those farms %ith the highest level of production build up any significant level of cash balance, while the rest, constituting the majority, do not have a cash surplus, not even with the loer level of home consumption. Farmers' attitudes Farmers were questioned as to constraints, their attitudes and ideals. Their answers are summarised below. With regard to farm inputs they revealed the following situation: ae Fertilizer: (60 percent of all farmers) 60 percent : costs are too high 21 " : not readily available 16 "' : transport not available 3 " : lack of know-how b. Improved seeds : (35 percent of all farmers) 71 percent : costs are too high 21 " : not available 8 " : transport not available c., Pesticides: (13 percent of all farmers) 50 percent : costs too high 20 " : not available 30 " : lack of know-how d. Tractor services: (19 percent of all farmers) 86 percent : costs too high 14 " : not available - 20 - TABLE 2.5: (a) Farm and Famnily Income, Home Consumtion and Household Expenditures, Kilimanjaro (shs) Details of income and _ Level of production expenditure Very Very Average high High Medium Low low Gross margin (farm income) 18,203 7,361 5,078 3,646 2,076 7,129 Off-farm income 670 874 540 1,108 1,608 952 FamilZ income 18,873 8,235 -5,618 4,754 3,684 8,081 Home consumption 10,178 4,260 3,206 2,375 1,376 4,199 Household expenditures 3,765 3,650 2,840 3,017 3,013 3,747 Cash for investment, replacement, savings, 4,930 325 -428 -638 -605 135 consumption. (b) Composition of Household Expenditures Item Shs Percentage of Total Clothing 1,599 42.7 Alcoholic drinks 717 19.1 Cooking oil 448 11.9 Sugar 381 10.2 Stimulants 301 8.0 Tobacco 280 7.5 Tea 3 0.1 Others 18 0.4 Kerosin 153 4.1 Condiments 56 1.5 Transport 22 0M6 Others 70 1.9 TOTAL 3,747 100.0 - 21 - Changes in farm organization d.ua to inqc'rease/decrease in producer price for various yroducts (surprisingly yielded the follwPing): a. Coffee (88 percent of all farmers): 88 percent anticipated a fluctuation f n culQ:'0yne are-a acccz line tc Troduct 12 percent anticipated no changeprclels b. Mai ze (93 percent of all farmers): 16 percent :fluctuation with the pri-e 4 " : constant acreages 80 ~r : increase in areas with increase in price; otherwise constant.0 Potential enterprises and reasons for their not being undertaken at present: - Wheat (8 percent of all farmers): 100 percent: unavailability of land0 - Pyrethrum (4 percent of all farmers): 100 percent: unavailability of land. - Cattle (8 percent of all farmers): 50 percent : investment costs high 33 : diseases 17 " : unavailability of grazing land. - Sheep, goats, pigs (8 percent of all farmers) : 50 percent :investment costs too high. 33 percent ;market for pigs not available. 17 percent: disease. - Poultry (7 percent of all farmers): 100 percent :disease All constraints removed, farmers would re-organize their farms as follows: Coffee/banana 0.99 ha Maize/beans 1.45 ha Pyrethruim 0.18 ha Wheat 0.06 ha Rice 0.06 ha Bananas 0.04 ha Orchards 0.03 ha 2.81 ha 22 - Apart from being ambitious as far as total cultivation size is concerned, farmers indicated increases in acreages for those crops for which the foregoing analysis revealed a need for application of more labour arl ca-ital per hectare (maize/beans). This, however, could easily be explained by the coiz:, 1--nn land being relatively non-expandable, and the cultivation of other crops constituting the only means for expansion, not or'.y in farm area but also in farming returns. In addition to crop enterprises, farmers would keep livestock in the following proportions: Type of livestock Percent of Head of livestock per farmers livestock keeping farmer Cattle 74 396 Sheep 21 6.3 Poultry 3 17.5 Pigs 3 6 Goats 3 3 The way farmers would utilise additional farm income is indicated by the following responses: Preferences, Dercent 1st 2nd 3rd Housing 57 22 11 Investment in agriculture 34 31 - Additional working capital 1 2 5 Education of children 4 26 23 Consumption 4 17 58 Savings - 2 3 TOTAL 100 100 100 Thus, farmers would first and foremost utilise additional incotes for the construction of new/better homes and invest in agriculture (livestock, land, tractors). Preference is shifted to education and increase in consumption, folloving their second choice. Consumption is increased as a third choice, indicating that farmers have room for better and expanded consumption, but with first priority to improve their housing situation (which includes to a certain extent questions of prestige) ane to expand and improve their farming operations. - 23 - Farm Programming In Kilimanjaro area two main crop mixes are all the crop activities there are. For farm programming both their average and-best performance coefficients are used. These data are available from Tables 2.2, 2.3 and appendix to Table 2.1. Both gross margin and variable costs are adjusted to exclude seasonally hired labour costs. The objective function initially is formulated to maximise gross margins with present resources, land, labour and variable costs held constant. As expected due to itscdefinite superiority, the coffee-banana mixture should be expanded to 0.93 ha, by using the technology and methods as presently applied by 20 percent of the top farmers. In addition, cattle keeping should remain an integral part of farming in Kilimanjaro area. Because present resources would be used up for more intensive coffee-banana cultivation, the mixture maize-beans would only be grown on 0.05 ha. 0.20 ha of farm land remains unused. The sum of gross margins ftim both cattle keeping and crop production in above proportions increases to 10,984 shs for the average farm, which is 54 percent more than at present. -According to this LP solution, coffee-bananas are grown un larger acreages than might be available in reality. Restricting this mixture to its present cultivation size of 0.77 ha, again the technology level of best farmers should be applied. In addition, livestock remains as at present and maize-beans cultivation should cover 0.25 ha using average technology. 0.16 ha are unused. Constraints are still variable costs which are subsequently increased by 50 percent to 886 shs/farm, Such a change, which could be accomplished by such measures as the introduction of small farmer credit schemes, would result in a change in cropping patterns as follows: Coffe-bananas 0.77 ha (best technology) Maize-beans 0.25 ha average technology) Maize-beans 0.16 ha (best technology Livestock as present with subsequent change in the sum of gross margins to 10,761 shs, which is almost as much as in the above case where the coffee-banana acreage was not restricted to its present level. The amount of variable cost still is the major constraint and even higher levels than assumdd could profitably be used. The above cropping pattern could be accomplished by the farm family without the help of hired labour. Family labour use would increase from 33 to 52 percent of its potential. Finally, the coffee-banana area is restricted to present levels and the level of variable costs are left open. Results indicate that, - 24 - due to an increase of the latter to 1,247 she, which is two times present average luvels, maize-beans could be grcun on 0.41 ha at the best technology level, in addition to the coffee-* anana. Labour is not a constraint in any month. Family labour use increases to 57 percent. The sum of gross margins increases to 11,796 shs, which is 65 percent higher than present, the average. In summary, Kilimanjaro farms can substantially be improved by spreading the technology of the best farmers into both coffee-banana and maize-beans cultivation to the mass of farmers. In addition to the transfer of better technology, substantial increases in additional working capital (660 shs/farm) are required. This could be accomplished by provision of small farmer credit, which, however, seems only realistic if at the same time the supply conditions of agricultural inputs (fertilizers, pesticides, seed, etc.) is improved. Further possibilities for improvement of the Kilimanjaro farms relate to expansion of the maize-beans acreage, through the use of animal and/or mechanical power, a case which is not tested by the present study. 2.1.2 Maize-Beans-Vegetable Farming in Lushoto The Lushoto survey area in the Usambara Mountains covers the divisional census areas 1621, 1631, 1633-35, 1637, 1641, 1643, 1663, 1665-6. This area was previously surveyed in 1965-1966 and comparisons will be made between results of the two surveys where appropriate. Attems 1/ describes the area as having a rather favourable climate, soils which are mostly poor in nutrients, and relatively good access to well supported markets. He states further: "there is hardly any other district in East Africa where development efforts began as early ....., were so manifold and so often repeated. In spite of this, however, the indigenous agriculture experienced a process of involution instead of evolution"a In the light of this assessment of the past, the results of this study and its comparison vith the 1965-66 survey are interesting. Summary Table 2.6 presents the results of 62 survey farmers, their parameters being also sub-grouped according to level of farm production. 1 ATTEMS, M[. "Permanent Cropping in the Usambara Mts," in Ruthenberg, H. Smallholder Farming and Smallholder Development in Tanzania; Mnchen 1968. -25- The Lushoto farm family consists of 5.6 people, forming a work force of 2.78 man- equivalents (6 and 23 respectively for .the 1965-66 survey)'. Divergenoies to these averages are notable only for the best farmers, who besides higher education and more people also have a larger work force. Their holdings are 24 times larger than the average, with cultivation size of (all by hoe) 1.09 hectares (of 0.97 hectares plus 0.08 hectares fallow land in 1965-66). A mixture of maize-beans occupies 50 percent of the cultivated area, maize in purestand 21 percent, coffee-banana 22 percent, and vegetables 4.5 percent. Differences between farm groups relate to a substitution of pure stand maize for the maize-bean mixture as moves towards low productivity farms. (Differences in the 1965-66 survey are that coffee-banana held a larger proportion and maize a much lower proportion in the average cropping pattern.) Total farm production is valued at 2,007 shs per farm on average. However, this average is only reached and surpassed by the 20 percent best farmers, who realized 6,762 shs on average. Differences in absolute level of farm production, of course, are to a large extent due to difference in farm sizes, but also to r.anagerial ability (e.g. rise in overall productivity by better resource combination and more optimal cropping patterns). Thus the mixture of maize-beans contributes more to the farm production than its proportions in the cropping pattern, and maize less, Variable costs amount to 189 shs on average, but are similarly low for 80 percent of all farms and are only slightly higher for the upper 20 percent best farmers (comparative data from 1965-66 relate only to purchased inputs). Labour inputs per farm average 1,706 hours, and are doubled in the case of best farmers. (In 1965-66 198 man-days were calculated, which, assuming 6 hour day, gives a lower figure.) The average farm income is 1,818 shs, with a ranFe of 297 to 6,057 shs between the poorest and best farm groups. These differences are, to a large extent, compensated for by off-farm income which roughly follows the opposite trend and increases from the best to the poorest farm groups. Thus family incomes average 4,147 shs ranging from 2,626 to 6,532 shs. (Off-farm income averaged 385 shs only in 1965-66.) Deduction of the value of home consumed products and household expenditures, which averaged 968 and 1,442 shs respectively, leaves an amount which is at the farm family's disposal for investment, replacement of capital goods, savings or additional consumption. 26 - SU1IARY TABE 2.6: Conosite Characteristics of Lushoto Farms - grouped by level of productivity - Level of production Average Datail sAvrg F-Very I Very _Hhigh IHh Medium Low low No, of ob-er;ati.ons 12 12 12 13 13 62 r-r, famriy' 1bour 45-8 52.0 36.8 42.7 48.9 45.2 IFariwer's zcation, school- ycarz, 2.3 0.5 1.2 0.6 0.7 1.0 I-rm fcnily, poople 7.4 5.8 4,7: 5.1 5.3 5.6 Parm, hirod labour, nos, - Total people on farm 7-4 5.8 4.7 5.1 5.3 5.6 Labo-r availability man- - ccuivalent s 3.58 2.94 1.90 2.71 2.94 2.78 Farm size, ha 2-57 1.24 0.77 0.55 0.43 1.09 (standard error) (0.44) (0.14) (0.17) (0.14) (0.11) (044) Cropping pattern, fo J.ais2-beans 72.4 34-7 37.7 20.0 18.6 49.5 Haize 6.2 23.4 31.2 32.7 60.5 21.1 Coffee-banana 14.4 28.2 26.0 40.0 18.6 22.0 Vegetables 2.8 5.6 3.9 3.6 2.3 4.5 Others 2.8 8.1 1.2 3.7 - 2.9 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 Livestock, head of cattle 2.58 1.25 1. "5 0.54 c.15 2.2 Variabole inputs, shs 706 91 92 40 38 189 Percentage distribution: Seed 98.5 100.0 100.0 100.0 100.0 99.3 Fertilizer 1.5 - - - -0.7 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 Labour innuts, hrs (not costed) 3,433 2,048 1,815 1,039 375 1,706 Percentage distribution 42.1 30.3 24.0 28.6 40.2 39.3 Livestock production 57-9 69.7 76.0 71.4 59.8 60.7 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 Value of farm production, shs 6,762 1,615 948 628 335 2,007 (standard error) (1,265.62 (70.84) (34,48. (24.42) (35.02) (399.67) Percentage distribution: Maize-beans 86.4 36.6 21.7 28.5 33.9 67.6 Maize 1.4 9.4 14.5 18.0 56.5 6.8 Coffee-banana 2.8 36.8 25.2 25.2 3.1 11.6 Vegetables 2.3 4.8 3.5 2.4 6.2 2.9 Others 1.1 3,6 6.2 1.9 1.3 2.0 Livestock 6.0 88 23.9 24.0 - - TOTAL 100.0 100.0 10 1001000 100.0 100.0 - 27 - IUM11ARY TABLE 2.6: (cont'd) Det ai Level of farm production Average Very Very high High Medium Low lo_T., ;Value of farm prod. shs 6,762 1,615 9./8 628 335 2,007 Variable costs, shs 706 91 92 40 38 189 Sum of gross margin, sh8 (farm income) 6,056 1,524 856 588 297 1,818 Off-farm income, shs 475 . 3,970 1,770 2,097 3,333 2,329 Family income, shs 6,532 5,494 2,626 2,685- 3,630 4,147 Value of home consumption, shs 2,860 752 433 524 271 968 Household expenditures, shs 2,290 1,358 1,260 1,293 1,051 1,442 Cash for investment, replacement, savings, consumption 1,382 3,384 933 868 2,308 1,737 LAND PRODUCTIVITY Value of production, shs/ha 2,631 1,302 1,231 1,141 779 1,841 :Variable costs, shs/ha 275 73 119 73 88 173 ,Gross margin, shs/ha 2,356 1,229 1,112 1,068 691 1,668 Labour input, hrs/ha 562 500 565 540 351 614 LABOUR PRODUCTIVITY Value added, shs/man-equiv. 1,692 518 450 217 101 654 Value added, shs/labour hrs used 1.76 074 0.47 0.56 0.79 1.06 Ratios Value of output to value of input 9.6 17.7 10.3 15.7 8.8 10.6 a. Labour not costed 9.6 17.7 10.3 15.7 8.8 10.6 b. Labour costed 1.7 0.7 0.5 0.6 0.8 1.0 Degree of family labour use, fo 64 46 64 26 8 41 Degree of commercialisation,% 58 52 53 17 19 52 Farm income as percent of total income 93 28 32 22 8 44 Cash balance as percent of total income 21 61 35 32 63 42 - 28 - This amounts to 1,737 shs on average and represents between 21 and 63 percent of total family income. This figure should, however, not detract from the fact that if farm income alone is taken as the basis for .comparison, no cash'balance would accrue for 80 percent of all farmers. Off-farm income is imnortant in Lushoto and accounts for 56 percent of total incomes on average. For the bulk of farmers, off-farm income amounts to as much as 68 to 92 percent of total incomes. The Lushoto Farm The Lushoto farmer is 45 years old, has an average of one year's schooling and is head of 4.6 family members Total work force amounts to 2.78 man-equivalents. 16 percent have additional employment to farming and earn, for an input of 250 days, 12 shs/day. On average this would amount to 484 shs/year. Nzo ujamaa activity was recorded from the sample used. The Lushotofarm, on average, comprises 1.09 hectares of land. The farm size structure is as follows: Farm size group, ha Average farm Percentage of Percentage of size, ha farms total area Under 0-50 ha 0.24 35-5 7,7 0-50 - 0.99 ha 0.74 24.2 16.4 1.00 - 1.49 ha 1.28 14.5 17.0 1.50 - 1.99 ha 1.75 12.9 20.6 2.00 - 2.49 ha 2.04 4.8 9.1 2.50 - and over 3.94 8.1 29.2 1.09 100.0 100.0 The structural distribution indicates a significant inequality in farm land availability. 35 percent of all farms cultivate only 7.7 percent of the total farm land, while 8 percent (larger farms) have access to more than 29 percent of the total farm land in the area. - 29 - The cropping pattern shows a variety oficrops cultivated, although only few have significant importance. These are: Cultivation Percentage on Crop or crop mix size, ha. cultivation size Maize/bean mixture 0.54 49.5 Maize 0.23 21.1 0offee/banana 0.15 22.0 Vegetables 0.06 4.5 Other crops (bananas, teal cardomom, cassava, yams) 1.09 100.0 42 farms or 72 percent in addition to crop cultivation keep cattle, on an average 3 head, or an average over all farmers of 2.2 head. Maize-bean 72 percent of all farmers grow a mixture of maize-beans on 0.71 ha. on average. Inputs, as well as outputs, of both components of the mixture cannot be separated. Productivity in total is high, with 2,421 shs/ha, ranging from 639 to 6,882 shs/ha, when stratified over 5 equal productivity groups. Variable costs average 307 shs and follow the almost linear trend of the value of production. Labour inputs for land preparation, planting and weed control, follow a similar pattern. Cultivation sizes are slightly smaller in the highest productivity groups. The highest gross-margins are three times higher than the average and more than ten times higher than those achieved by the group of poorest performance (See table 2.7). Compared to the same crop .,ix of Kilimanjaro area, which is operated not too differently, the results are very similar. The difference relates to much higher levels of variable inputs in the latter, which compensate for small inputs of land. A Cobb-Douglas production function analysis produced the following production elasticities: land 0.01; labour in land preparation 0.17; labour in weed control 0.14; variable inputs 0,71*. The sum of elasticities is 1.03 and the adjusted R2 0.84. Applied to average products and inputs, these elasticities result in marginal products much higher than likely opportunity costs for variable inputs (seed), and labour for all operations: variable inputs 5.62 shs/ha, labour 2.52 shs/hr, and 2.38 shs/hr for land preparation and weed control respectively. - 30 - TABLE 2.7: Input/Output Coefficients b;y Level of Prod.uctivity - Maize-Beans, Lushoto - Level_of productivity Input/output coefficients I Average Very~ Very high High Medium Low low No. of observations 9 9 9 9 9 45 Cultivation size, ha Co.50 0.60 0.96 0.77 0.77 0.71 Variable inputs, shs/ha Labour, land prep. hrs/ha 612 397 206 198 1 139 178 weeding, it 567 332 172 166 118 143 others, 453 162 129 113 77 103 TOTAL 1,614 991 507 447 334 424 Value of production, shs/ha 6,882 3,287 27236 1,326 639 2,421 Variable costs, shs/ha 758 285 317 235 84 307 Gross margin, shs/ha 6,124 3,002 9 11,091 555 2,114 Marginal products for land amount only to shs 24/ha against which opportunity costs can be measured. The overall indication is that the mixture of maize and beans should, as for the first survey area, be more intensively cultivated with additional labour and much more capital. An adjustment towards more optimum combinations of inputs indicates a rise in farming returns to 2,333 shs. This input combination includes the use of 83 labour hours in land preparation, 68 hours in weed control and 348 shs variable costs on 0.05 ha of land. These results are only indicative, pointing towards the course of action for achieving higher returns for this miture. Maize Maize in purestand is grown by 55 percent of all farmers on 0.38 hectares on average. Yields average 505 kg/ha. Highest productivity levels reach 2,333 kg/ha. This is achieved by 20 percent of all maize growingfarmers with distinctly higher material and labour inputs per hectare and on much smaller cultivation sizes (Table 2.8). - 31 - TABLE 2.8 Inpgt/Output Coefficientseby Level of Productivity - Haize, Lushoto Input/output coefficients Level of productivity Average N I ~ -A Very Very --__ . high High Medium Low Low I I i I ITo. of observations It 6 7 --_- 7 7 34 Ci.ltivation size, ha 022 2 0. - 0.48 0.54 0.38 Va.-ri able inmut s, shs/ha Labour, Land prep, hrs/ha 567 160 138 111 160 146 Weed control, hrs/ha 473 165 125 85 128 125 OtharE, hrs/ha 228 72 51 47 47 53 TOTAL 1 ,26- A 397 314 243 335 324 Yield kg/ha 2,333 670 496 387 282 505 Value o rduction shs/ha Variable costs, shs/ha 260 98 58 59 52 .66 Gross margin, shs/ha 2,677 783 578 446 316 569 Production function analysis produced, as for the mixture of maize-beans, high elasticities mainly for -ariable costs (seed) and labour for land preparation, viz:0,43* and 0.23* respectively. Both land and weeding labour each yielded an elasticity of 0.05,trinnthe sum to 0.76. The adjusted R2 was only 0.34. Marginal productivities have been calculated using three levels of producer prices: present average, 25 percent lower than average and 25 percent higher. Since the proportions of marginal products of the various factors do not change, also their optimal combination stays the same. Thus, taking the present total costs as given, the optimal combination is as follows: 0.11 ha of land, 48 labour hrs in land preparation, 11 in weed control, and 91 shs variable costs (seed). Returns ame increased by almost 100 percent to 1,137 shs on average, and to 853 shs under the low producer price assumption and 1,421 shs under the higher producer price assumption. Productivity increases even more. Thus maize should be cultivated much more intensively, applying higher rates of labour and capital inputs. - 32 - Coffee-banana 24 percent of all farmers grow a mix-ture of coffee-bananas on average of 0.63 hectares. Value of prodction amo',is to 1,506 shs/hectare, which is less than in the previous survey area. Lalbour f.:yes average 837 hours/ha, of which more than half is for weed control (Table 2.9). Proation function analysis yielded inconclusive results. Coffee in purestand is grown by 14 percent of farmers on 0.43 hectares each. The productivity is less than for coffee grown in mixture with bananas. Also labour inputs are lower in about the same proportion as productivity. Production function analysis did not produce result, which could be used for further analysis. 9 percent of farmers grow on average of 0.32 hectares bananas in purestand. Productivity is disappointingly low, amounting to only 289 shs/ha on average. This could be partly due to under-recording of the continuously forthcoming harvest of this crop. Vegetables Vegetables are cultivated by 14 percent of farmers on an average of 0.32 hectares. The value of production amounts to 1,305 shs/ha and the gross margin, after deduction of 175 shs/ha for variable costs, to 12130 sho (Table 2.9). Production function analysis indicated a negative elasticity for land (-0.43), and positive elasticities for variable inputs (0.14), weeding labour (0.78) and cultivation labour (0.35). The sum of elasticities is 0.84 and the adjusted Z2 was 0.61. Although optimal resource combination has not - due to negative ratio for land - been attempted, it seems possible to raise returns from vegetables through a much more labour intensive cultivation, especially in weed control. Cassava 4 farmers grew cassava on 0.3 hectares each. Value of production seems to be under estimated, due to continuing harvesting practices, which are difficult to recall. This seems also to be the case for the labour records for the same operation. (Table 2.9) - 33 - TABLE 2.9: Input/Output Coefficients for Various Crops - Lushoto Input/output Annual crops Perennial crops coefficients Cassava Vege- Bananas Coffee Coffee/ Cardomom Tea tables banana No. of observations 4 9 9 5 21 Cultivation size, ha 0.30 0.32 0.32 0-43 0.65 0.58 0.75 No. of plants/ha 1,183 931 1,336* 1,453 4,000 Establishment of I costs, shs/ha 1,332 1,656 1,229 1,475 2,000 Prod. lifespan, years 20 44 45 7 55 kA-nual depreciation, shs/ha 66 73 25 211 36 Variable inputs, shs/hI Seed 14 175 - - 7 30 - Fertilizer - - - - - - 107 TOTAL 14 175 -7 30 107 Labour, hrs/ha Land prep/planting 242 242 - - 18 55 - Weed control 311 193 223 344 457 82 96 HarTesting 93 -145 140 365 362 248 13 TOTAL 645 587 363 709 837 385 1,429 Yield, kg/ha - - - 183 - 262 - Value of prod. shs/ha 468 1,305 289 1 1,257 1,506 1,034 960 Gross margn, shs/ha 454 1 1,130 289 1,257 1,499 1,004 853 * Coffee only Cardomon This crop was cultivated by 2 farmers on 058 hectares on average. Yield amounted to 262 kg/ha valued at 1,034 sho. Labour amounted to 385 hours/ha of which more than 60 percent was for harvesting. At 1,475 shs estimated establishment costs and a productive life of 7 years, yearly overhead costs amount to shs 211/ha. (Table 2.9) Tea Tea was recorded to have been cultivated on 0.75 ha by one farmer. The gross margin of 853 shs has to be seen against the very high labour input of 1,429 hrs/ha and annual overhead costs of 36 shs/ha originating from establishment costs of 2,000 shs, depreciated over 55 years. (Table 2.9) - 34 - Livestock More than 70 percent of all Lushoto farmers keep livestock in the form of cattle, with an average of 3 head/farm. Labour inputs are estimated at 2,696 hours/year and value of production at 437 shs, including 269 shs for inventory changes. Cobb-Douglas production function produced similar results as for the Kilimanjaro area: an extreme low R 2, low sum of elasticities and extremely low marginal-returns-to- opportunity-cost-ratios for the only two quantifiable variables, labour and head of cattle. Whole farm On average, the value of farm production amounts to 2,007 shs, contributed by 1.09 hectares of cultivation and 1.13 head of cattle. The sum of gross margins amount to 1,818 shs. This result is achieved by an average labour input of 1,706 hours, about 60 percent of which is used for cattle keeping (Table 2.10). Farming returns range from 335 shs to 6,763 shs/farm. Outstanding results are achilved by 20 percent of all farmers. These apply higher levels of variable inputs including labour, the latter mainly used for crop production. Farms with the highest returns are more than twice the size of the average and have more cattle per farm. The same major trends are evident, when farms are grouped by farm size. Farming returns increase with increase in farm size, although rather moderately and almost linearly until the last group where cultivation size is more than three times the average, and values of production are 4- times larger than average. The issue .is further explored by Cobb-Douglas production function analysis. All variables included shou positive elasticities, which are: land 0.07; variable inputs 0.38; labour (crops) 0.52*; labour (livestock) 0.09; livestock numbers 0.06. Their sum is greater than one, implying increasing returns to scale. The adjusted R2 is 0.81 Farming returns increase as more of all resources are used (with their proportion held constant). Only labour for livestock and the number of animals have marginal returns to opportunity cost ratios much less than unityp Use of all other inputs should be increased, especially variable inputs. - 35 - TABLE 2.10 (a): Input/Output Coefficiontc fo' Farms by Level of Value of 7?-o lu-tion - Lush:to f 19ve of production Input/output coefficients -I-- --'_production Average chry Very hieh High__ Medium Low i low No. of observations 12 12 12 13 13 62 Cultivation size, ha 2 1-24 0.77 0.55 0.4 1.09 Head of cattle, no,. ,2_8 1 ' 1.25 0.54 0.15 1.13 Labour input, crops, hrs 1,445 621 436 297 151 670 Livestock, hrs 1,988 1,427 1,379 742 224 1,036 TOTAL, hrs 3,433 290A8 1 81 1,039 3 1706 Value of production,shs 6,763 1,615 948 628 335 2,007 Variable costs, shs 706 91 92 40 38 .. 189 8567 1 88 29 Gross margin, shs 67057 1,524 856 5297 1,818 (b): Input/Output Coefficinr'.o for Farms by Farm Size Classes - Lushoto - Farr size group ha Input/output coefficients 7-r .50 - 1.00 - 1.50 - 2.00 - 2.50 & _-50 .99 1.49 1.99 2.49 over No. of observations 22 15 9 8 3 5 Average farm size, ha 0.24 0.74 1.28 1.75 2.04 3.94 Value of production, shs 563 1,328 1,501 2,957 2,760 9,336 Variable costs, shs 51 93 156 270 78 1,072 Gross margin, shs 512 1,235 1,345 2,687 2,682 8,264 Labour, crops, hrs 202 447 943 847 1,047 1,252 Livestock, hrs 862 1,327 486 2,232 - 1,798 Total, hrs 1,06z4 1,774 1,429 3,594 1,047 3vo50 Head of cattle no. 0.73 1.07 0.78 2.0 0 3.0 _ _ _ _ _ _ _ _ _ _ _ _ _ I _ _ _ _ _ _ _ _ _ _ _ _ __J - 36 - The optimal combination of resources which yields, at present cost levels, 5,244 shs requires farm sizes of 1.55 ha, variable inputs valued at 793 shs, labour inputs for crop cultivation of 1,144 hrs. Not only would these incr---- 1 s cut productivity would also be increased by 84 percent. Again these results are only indicative of the direction in which farms should develop for increased rtturns. Individual components of the farm business have individually to be adjusted. In Table 2.11, the income situation of the Lushoto farms is analysed. The farm income (sum of gross margins) averages 1,818 shs. Off-farm income in the form of remuneration, business profits and family contributions amounts to 2,329 shs/farm. Both sources add up to family income of 4,147 shs, from which home consumption of 968 sh;/farm and household expenditures of 1,442 shs/farm have to be met. This leaves the farm with a cash balance of 1,737 shs. However, this average should not detract from the fact that for 80 percent of farms farm incomes alone would not be sufficient to meet present household expenditures, let alone accrual of any cash balance for increased investment, replacement or savings. Table 2.11 also gives the breakdown of expenditure items. Almost 60 percent of total expenditures is used for household goods and clothing. Large amounts are also spent for sugar and cooking oil. Farmers' attitudes Farmers in Lushoto are especially concerned. with the lack of fallow land and the resulting low fertility of the soil. They also feel that prices for most of their -roducts are inadequate (72 percent of the farmers), especially for coffee and maize. Bananas are mentioned by 26 percent of the farmers as having a low demand. 52 percent of all Lushoto farmers consider the availability of material inputs as a serious constraint. This related especially to: (improved) seed : 65 percent insecticides : 25 " manure : 25 " Fertilizer is not mentioned in this consideration, but its price is considered by 74 percent of all farmers as excessively high, a reason why its use is not widespread. - 37 - TABLE 2.11: (a) Farm and Family Income, Home Consumption and Household Txenditures - Lushoto - Details of income and Levels of production A expenditure IVe ry Very I high High Medium Low low Gross margin (farm income) 6,057 1,524 856 588 297 1,818 Off-farm income 475 3,970 1,770 2,097 3,333 2,329 Family income 6,532 5,494 2,626 2,685 3,630 4,147 Home consumption 2,860 752 433 524 271 968 1Household expenditures. 2,290 1,358 1,260 1,293 1,051 1,442 Cash for investment, replacement, savings, 1,382 3,384 933 868 2,308 1,737 consumption. (b) Composition of Household Expenditures Item Shs Percentage of total Household goods 433 30.0 Clothing 415 28.8 Sugar 191 13.2 Cooking oil 186 12.9 Stimulants 62 4.3 Coffee 12 0.8 Tea 28 2.0 Tobacco 22 1.5 Detergents 62 4.3 Kerosine 60 4.2 Condiments 33 2.3 TOTAL 1,442 100.0 Farmers indicated that if constraints were removed, they would grow the following crops in the following proportions. Percent Maize 39.0 Beans 17.2 Tea 18.5 Coffee 13.5 Vegetables 7.4 Cardomom 4.4 100.0 - 38 - This would mean an increase in average farm size to 1.35 ha. This intended restructured cropping pattern contains more and diversified cash crops, especially tea. Almost all farmers (93 percent), in addition to crop enterprises, would like to keep cattle, on average 10 head per farm. Thus, there is a preference for mixed farming. Farmers' preferences in spending additional income would be as follows: Preference, percent 1st 2nd 3rd Construction of new house 56 22 10 Investment of agriculture 44 75 74 Consumption - 3 16 Thus, farmers would first and foremost improve their homes. Equally, they are keen to improve and invest in their agricultural business, mainly by buying both land and cattle. The percentages also include additional use of material inputs. Finally, consumption appears only as their second choice and even as a third choice it does not seem to be of much importance. Farm programming In farm programming both available activities and constraints are used. Activities of Lushoto area refer to the main crops and crop mixes as determined by the survey: maize-beans, maize, vegetables, coffee-bananas, and coffee. For all of these the gross margin, seasonal labour requirements and variable costs are provided in the foregoing. Constraints refer to the farm size cultivated (1.09 ha), the labour availability (2.78 man-equivalents, which provide 348 hours of labour per month) and the working capital averaging 189 shs/farm. Also introducing the highest productivity levels of maize and maize-bean cultivation, whose gross margins are by far superior to those of average productivity levels of all crops, LP indicates as optimum solution the cultivation of 0.73 ha of maize, with-a gross margin of 1,946 shs per farm, which is slightly higher than present farm income (1,818 shs). As much as 0.36 ha of farm land is left unused, mainly because of constraints in the supply of working capital (present level of variable costs). This situation rectified, the farm organization would include: 0.74 ha maizerbeans (best productivity) 035 ha maize (average productivity) 0/0 - 39 - )nd 669 shs of working capital. This farm organization would yield a gross margin of 5,280 shs per farm on present average farm sizes. Labour.becomes constraining in the c:onth of August and does not allow livestock to be kept in addition to above crops. U2'er the assumption that additional labour could be hired during August, the above cropping pattern would change to include a maize-bean mix on the total available cultivation size of 1.09 ha, using the best productivity level available. In addition one head of cattle could be kept. Gross margin per farm - accounting for the costs of 200 seasonal hired labour hours during the month of August - amount to 6,673 shs/farm. This is more than three times present average. Working capital needs to be increased from 189 to 1,026 shs/farm, which again calls for the use of small farmer credit, accompanied by an adequate supply of agricultural inputs and the necessary supporting s3rvices in agricultural extension. The potentials are there to transform present small farmer agriculture into a viable one. 2.1.3. Maize, Coffee and Pyrethrum Farming in Mbeya The Mbeya survey area is delineated by the divisional census area 0831. Results are available for only 29 farmers and they are presented in Table 2.12. The Mbeya-farmer is 45 years old and has approximately one year of formal education. The farm family averages 5.2 people forming a po ential work force for agriculture of 3.6 man-equivalents. 1.68 hectares are cultivated, 40 percent in maize, 18 percent in coffee and bananas, 17 percent in pyrethrum, 15 percent in pulses and oil crops and the remainder in wheat and millet. The better farmers in Mbeya are older than the average and slightly larger families and work forces. They especially cultivate more pyrethrum and less wheat and millet. Their farm production is 2i times the average, the latter amounting to 4,342 shs/farm. The higher farm production is not solely due to larger cultivated sizes but also to increase in productivity and more optimal resource allocation to individual enterprises. Maize, pulses and pyrethrum contribute on average more to the farm production than their respective proportions in the cropping pattern, although the pattern shifts from one farm group to the other. Certain crops must have similar productivity levels and probably for this reason no definite trends are visible. - 40 Variable costs include seeds, fettilizers, pesticides and seasonally hired labour and average 564 shs/farm. Seasonal hired labour is mostly used in the higher farm groups and amounts to 44 percent of the total cost. Labour inputs average only 582 hours per farm. They are - although not proportional to the farm size - higher in the group of best performing farms. The sum of gross margins or farm income, ranges from 958 shs to 9,506 shs and averages 3,776 shs per farm. Sizeable off-farm incomes only show for the highest farm groups, thus idening the gap between this group and the rest of the farms. Family incomes amount to 4,621 shs per farm, from which home consumption and household expenditures of 1,444 and 865 shs per farm respectively have to be met. The averago farm is left with a sizeable amount of cash for investment, replacdment or saving, which is more than half of the total income. Remarkably, these cash balances are positive for all farm groups, which could be due to under-recording of expenditures, but probably also to the considerably lower levels of expenditure. Maize Maize in purestand is grown by 75 percent of all farmers. Some of these cultivate more than one field. Thus 30 observations become available and with an average cultivation size of 0.58 hectares. Table 2.13 summarizes the main features of maize farming in Mbeya. Maize yields average 2,275 kg/ha with best performing cases more than doubling this amount. This seems to be due to all the inputs being used in higher proportions than on average. Values of production, using the average producer price of 1.16shs/kg, amount to 2,639 shs/ha and gross margins 2,328 shs. Labour inputs are 363 hrs/ha. Cobb-Douglas production function analysis yielded high elasticities for both land (0.40*) and labour in land preparation (0.50*). Seed and pesticides have negative elasticities of -0.08 and 0,14 respectively. The elasticity for fertilizer was 0.12 and that of labour in weed control 0.05. The sum of elasticities was 0.85 and the 2 adjusted R 0.72. Marginal productivity for all factors with positive elasticities is higher than factor costs. This is especially so for land and labour in land preparation. This result confirms that maize grown on larger hectarage and with more labour has much higher potentialsthan at present. Both seed and pesticides are used at levels uhere additional inputs (at present application methods) are not expected to contribute to increased output. - 41 - SUMARY TABLE 2.12: Composite Characteristics of Mbeya Farms - grouped by -level of productivity - Level of production Detail Average Very Very -high High Medium Low low No. of observations 5 6 6 6 6 29 Farmer, family, labour Farmer's age, years 54.2 54.0 34.0 41.0 42.3 44.8 Farmer's education, school yrs 0.4 1:5 1.0 0,8 0.7 0.9 Farm family, people 6.0 6*7 4.5 4.7 4.3 5.2 Perm. hired labourers -- - . Total people on the farm 6.0 6*7 4.5 4.7 4.3 5.2 Labour availability9 man-equivt 3.7 5.4 3.0 2.7 3.2 3.6 FARM LAND Farm size, ha 3.50 2.26 0.91 1.04 1.01 1.68 (standard error) 1 (0.41) (0.51) (0.19) (0.18) (0.50)i (0.25) Cropping pattern, % Wheat, millet 1.7 9.3 9.2 19.0 40.0 10.4 Maize 40.1 48.5 26 1 44.0 29.2 39.0 Beans, peas,groundnuts 12.4 7.8 43:7 3.0 10.7 15.2 Pyrethrum 30.0 7.4 8.4 17.0 6.1 17.1 Coffee, bananas 15.8 27.0 12.6 17.0 13.8 18.3 - TOTAL 100.0 100.0 100.0 1l00..0 100.0 100.0 Livestock, head of cattle 1.2 1.2 2.8 - 0.08 0.7 Variable inputs, shs 1,502 1,026 264 108 77 564 Percentage distribution: Seed 12.4 43.1 44.5 38.5 43.2 24.9 Fertilizer 34.6 47.2 50.5 1 51.0 32.2 40.1 Pesticides 5.6 4.6 3.6 9.4 - 5.2 Hired labour 44.4 3.2 24.6 27.5 Hired power - - - - . . Others 3.0 1.9 1.4 1.1 - 2.3 TOTAL 100.0 100.0 100.0 100.0 100.0, 100.0 Labour inputs, hrs (not costed). icrops only) _________ 928 1,117 412 314 198 582 Value of farm production, shs 11,008 5,636 3,116 2,028 1,035 4,342 (standard error) (1,032.44)(669.00) :(297.81) (171.8?), (118-52) (706.32) Percentage distribution: Wheat, millet .2 3.1 1.7 2.2 14.5 2.2 Maize 39.8 53.8 40.4 56.4 74.6 51.4 Beans, peas, groundnuts 14.8 14.8 31.5 1.8 4.4 17.8 Pyrethrum 21.7 15.1 20.4 32.8 5.0 23.0 Coffee- bananas 22.2 12.6 5.3 6.8 4.6 Others 1.3 0 .6 0.7 -1.5 1.0 TOTAL 100.0 100.0 100.0 100,0 i 100.0 100.0 - 42 - SU-MARY TABLE 2.12 (cont'd) Level of farm production FDetail Vr Vey Average ahigh High Medium Low low5 Value of farm production, shs 11,008 5,636 3,116 2,028 1,035 4,342 Variable costs, shs 1,502 1,026 264 108 77 564 Sun of gro mgin, shs (farm income) 92506 4,610 2,852 1,920 958 3,776 Off farm income, shs 3,800 250 500 - 166 845 Fhmily income, shs 13,306 4,860 3,352 1,970 1,124 4,621 Home consumption, shs 3,100 1,904 1,433 625 433 1,444 Household expenditures, shs 1,625 612 1,094 837 339 865 Cash for investment, replacement savingst_ consumption 8,581 2,344 825 458 352 2,312 LAND PRODUCTIVITY Value of production, shs/ha 3,145 2,493 3,424 1,950 1,024 2,584 Variable costs, shs/ha 429 454 290 104 76 336 Gross margin, shs/ha 2,716 2,039 3,134 1,846 948 2,248 Labour input, hrs/ha 265 494 453 302 196 346 LABOUR PRODUCTIVITY Value added, shs/man-equiv. 2,749 859 951 711 305 19091 Valuo added, shs/labour hrs used 10.96 4.16 4a9 6.11 4.93 6.75 Ratios Value of output to value of innut a. Labour not costed 7.3 5.5 11.8 18.8 13-4 7.7 b. Labour costed 4.5 2.6 4.6 4.8 3.8 3.8 Degree of family labour use, 7 14 9 8 4 Degree of commercialization, o 72 66 54 69 58 67 Farm income as percent of total income 71 95 85 loo 85 82 Cash balance as percent of total income 64 4o 25 23 31 50 64 46 /. - 43 - TABLE 2.13: Input/Output Coefficients by level of Productivity - Maize, Mbeya - Input/output coefficients Level of productit Average Very Very high High Medium Low low No. of observations 6 6 6 6 6 30 Cultivation size, ha .0.47 0.35 0.68 0.40 1.05 0.58 Variable inputs, shs/ha Seed 156 112 77 66 62 89 Fertilizer 270 194 202 104 133 177 Pesticides 21 - 14 1 8 4 9 Others 50 - 34 1 52 36 TOTAL 497 306 327 179 251 311 Labour input, hrs/ha Land prep planting 245 152 91 112 89 145 Weeding 187 110 59 51 45 82 Harvesting, others 221 202 73 179 81 146 TOTAL 653 464 223 342 215 363 eld, kg/ha 4,629 3,224 2,501 2,o44 1,121 2,275 Yied i 22 .Value of production, shs/ha 5,369 3,737 2,901 2,371 1,300 2,639 Variable costs, shs/ha 497 306 327 179 251 311 Gross margin, shs/ha 4,872 3,433 2,574 2,192 1,049 2,328 The optimal combination of inputs derived by using the above results is as follows: fertilizer worth 34 shs, 150 labour hours in land preparation and 15 in we.ed control, and cultivation size of 0.70 ha. The result is an output of 2,204 shs, which represents an increaso in productivity of 20 percent. Coffee Coffee in:purestand is grown by 9 farmers (30 percent) on average cultivation sizes of 0.54 ha. Farmers estimate the number of trees per hectare at 1,318 and the life of the plantations at 51 years, although at present plantations average only 9 years of age. Yieldsamount to 547 kg per hectare or 0.42 kg per tree. Per hectare values of production, at producer prices of 3.09 shs/kg amount to 1,692 shs. Variable costs, made up mainly of expenses for fertilizers, amount to 258 shs, leaving a gross margin of 1,434 shs. According to farmers' estimates of establishment costs, another 247 shs should be deducted for depreciation, leaving a net return of 1,187 shs/ha. -44- Cobb-Douglas function analysis indicates positive elasticities for all factors: land 0.47*, fertilizer 0.21, pesticides 0.14, weeding labour 0.16. They sum up to 0.98 and the adjusted. R2 is 0.65. Marginal productivities are high for all factors except labour. Those for land and pesticides are much higher than their assumed factor costs. Optimum re-allocation of resources, restraining total variable costs to present levels, requires both the doubling of the acreage and increased use of pesticides, while labour for weeding would have to be curtailed to a third of present levels. Production would raise by 70 percent. The restriction to present total cost levels, however, results in a loss of productivity in the order of 40 percent. The present combination of resources is more labour and capital intensive, while the optimum combination maximizes total returns under a predetermined (present) cost level. Coffee-Bananas Four farmers (14 percent) cultivate a mixture of coffee-banana, on 0.65 ha each. The number of coffee trees in the mixture equals that of the purestand coffees and so do the yields. Additional income is obtained from the second component of the mixture, bananas. Value of production amounts to 3,633 shs/ha and gross margin, after deduction of costs for fertilizer (133 shs), pesticides (52 shs) and other (47) to 3,401 shs/ha. Annual depreciation of the plantation - if considered - would amount to 465 shs, thus yielding a net return of 2,936 shs/ha which is 2 , times higher than for coffee in purestand. Bananas Three farmers (10 percent) grow bananas on 0.23 hectares each. Records available are not consistent and were not analysed. Pyrethrum Thirteen farmers grow this cash crop on 0.62 hectares on average. For 10 farmers complete records are available. Inputs to this crop relate to land and labour only, the latter amounting to 380 l"rs/ha. Yields average 915 kg/ha, which converts, with a producer price of 3.94 shs/kg, to a value of production of 3,600 shs. '45 - 45 - Cobb-Douglas production function analysis yields for both land and labour in land preparation high positive elasticities.of 0.34 and 0.65 respectively. The sum of elasticities amounts to 0.99 and the adjusted R2 is 0.94. For both above resources much higher marginal returns are calculated than would be their factor costs, indicating that substantial increases in returns could be forthcoming from an expanded use of these factors. Optimal resource allocation, at present total cost levels, is indicated as follows: average cultivation size should be decreased to 0.27 ha per farm and 209 labour hours should be employed on this hectarage for land preparation and planting. Yield could thus be raised three times. Of course, higher labour inputs for picking would also be necessary. These indications do not include any technological change, such as better varieties, fettilizer use, etc. Wheat Wheat is grown on 16 farms on 0.23 hectares each. However, complete records are available from only 12 farmers, who, with an input of 60 shs worth of sedd and 206 labour hours, harvested 428 kg/ha. At average producer prices of 0.91 shs/kg, value of production amounts to 391 chs/ha and gross margin to 331. Cobb-Douglas production function analysis gave high positive elasticities for land (0.30), variable inputs (0.31*) and labour for land preparation (0.43*). Labour for weed control yielded only 0.03. The sum of elasticities is 1.07, and the adjusted R2 is 0.80. Marginal returns are much higher than opportunity costs for the variable input (seed) and labour for land preparation. Land and labour for weed control have ratios smaller than unity. Re-allocation of resources for maximizing returns emohasizes more intensive use of above resourc_-s: 24 shs worth of seed and 36 labour hrs in land preparation on a cultivation size of 0.15 ha, yielding, at present total cost levels, returns of 121 shs, which are 36 percent higher than at present. Also productivity increases by 100 percent due to the re-allocation of variable inputs of 166 shs/ha and labour inputs for land preparation of 264 hrs/ha. - 46 - Minor crops In addition to the above, the following crops and crop mixtures are grown in the area but on too limited a scale to permit detailed and meaningful analysis: millet, beans, peas, maize-beans, and maize-groundnuts. Livestock Five farmers, or 12 percent, keep cattle; an average of 7 head (ranging from 2 to 15), valued at 434 shs each. Altogether, 7 farmers keep goats and/or sheep, on an average of 5 sheep and 3.7 goats, valued at 57 shs per animal. Fourteen farmers, or almost half of Mbeya farmers, keep poultry,on an average 8 birds (ranging from 3 to 33 birds), valued at 9 shs each. - Information about inventory changes, production, labour and material inputs is not available to extend analysis of livestock keeping further. Whole farm The average Mbeya farm of 1.68 hectares produces a value of production of 4,342 shs. Variable costs amount to 564 shs and the sui of gross margins to 3,776 shs. Labour inputs are only 582 hours per farm (Table 2.13), livestock activities not included. The 20 percent of farmers with highest levels of production have all of these parameters more than doubled. Measured on the average productivity, which is 2,584 shs/ha, 82 percent of the increase in production of the best farmers is due to increase in scale of operation while 18 percent is due to increases in productivity. The increases in the latter, if applied to average farms could raise production to 5,124 shs/farm, if no such constraints as the need to ensure home consumption of certain staple foods are present. Table 2.13 also summarizes the major parameters by farm size groups. Smaller farms, although having absolutely smaller values of production, achieve substantially higher levels of productivity. 0ob-Dou,1as pr?od0mi.sn _ nct:n analys!is applied to whole farm data of Mbeya included a ta--: o - to total. fa. ize into major or.op-enterprises. Land devoted to wheat, millet: cof:'re-bincna. y-ieleftd negative elasticities and thus negative mar:.nal reurns. L-..- :'ethr:. ar' pulses - groundnuts, show high positive elasticities n- d: 1 : -::c:1 ear le scts. All elasticities are summarized below: -0.11 Tcnd, r. a 0.03 La=c, pls s, grounauts 0 20* Lect,gyrotrura0.20 LE_nd, cof:'ee, banana -0.02 LivToc'- -0.02 0.48* VriaOle costs 0.43* lum of elasticities 1.19 2 (aijusteO) 0.78 Opti-m-u re-allocation of resources, keeping costs restrained at present levels indicates cultivation sizes of 1.19 for both pyrethrum and pulses/groundnuts. In addi-tion 0.18 ha maize is grown. With 496 labour hours and 418 shs variable costs a gross margin of 10,889 shs is calculated, which is 2- times the return of the average farm. Farm size would be increased by 50 percent, but productivity would increase by 90 percent. In Table 2.15 the income situation of the Mbeya farmer is analysed. Farm income averages 3,776 shs/farm, with more than 60 percent of farmers obtaining incomes below this, and best performing farmers obtaining 2* times the average. The latter also are the only group which receive substantial income from other sources than farming. In this way the gap in total or family incomes is widened. Family income averages 4,621 she. In the highest farm group it amounts to 13,306 shs/farm. The value of home consumption amounts to 1,444 shs and household expenditures to 865 shs. Both figures gradually diminish from the highest to the lowest farm groups, Household expenditures are mainly for clothing (44 percent) and sugar (15 percent). The rest is divided between other consumer items, such as alcholic drinks, condiments, kerosine, cooking oil, detergents and stimulants. The cash available for investment, replacement of capital items, saving or additional consumption amounts to 2,312 shs or 50 percent of total family income. Even the lower farm groups have between 23 and 31 percent. These are surprisingly high figures and show - apart from possible occurence of over or under recording, a high inivestment potential for 11beya farmers, but possibly also curtailed consumption behaviour - 48 - TABLE 2.14: (a) Input/Output Coefficients for Farms by Level of of Farm Production - Whole farm, Mbeya - Level of farm production Very Avera-e1 Input/output coefficients Avera SVery Very high High Medium Low low No. of observations 6 6 6 6 29 Cultivation size, ha 3.50 2.26 0.91 1.04. 1.01 1.68 Cattle, head 1.2 1.2 2.8 - 0 ,08 0.7 Labour input, hrs 928 1,117 412 314 198 582 Value of production, shs 11,008 5,636 3,116 2,028 1,035 4,342 Variable inputs, shs 1,502 1,026 264 108 77 564 Gross margin, shs 9,506 4,610 2,852 1,920 958 3,776 (b) Input/Output Coefficients for Farms Grouped by Farm Size Farm size group, ha Input/output coefficients tnder 0.50- 1.00 1.50 - 2.00 - 2.50 & -- 1 0.50 0.999.49 1.99 I2.49 Labove No. of observations 5 7 7 2 2 6 Cultivation size. ha0.40 0.65 1.28 1.60 2.20 3.72 Cattle, head 3.00 0.62 1.00 ' 1.00 - 0.85 Labour inputs,.hrs 194-1 317 380 460 885 1,218 Value of production, shs 1,704 1,739 3,356 3,411 7,390 8,764 Variable costs, shs 149 100 215 289 1,168 1,503 Gross margin, shs 111639 31 6222 7,261 TABLE 2.15: (a) FPrm FeilyIncome 50 sumotion and Household Expenditure - I`be3rz shs) Detail of income and avel of production - - - IAverage expenditures Very Very h H-h Medium L Low low Gross margin (farm income) 9,506 4,610 2,852 1,920 958 3,776 Off-farm income 3,800 250 500 - 166 845 flmily income 13,306 4,860 3,352 1,970 1,124 4,621 Home consumption 3,100 1,904 1,433 625 433 1,444 Household expenditures 1,625 612 1,094 837 339 865 Cash for investment, savings, replacement, consumption 8,581 2,344 825 458 352 2,312 (b) Composition )f Household Expenditures Item Shh [ Percentage of total Clothing 384 44.4 Sugar 128 14.8 Alcoholic drinks 77 8.9 Condiments 76 8.8 Kerosine 50 5.8 Cooking oil '0 4.6 Detergents 44 5-1 Stimulants 37 4.3 Tea 26 3.0 Tobacco 11 1.3 Others 29 3.3 TOTAL 865 100.0 0/0 - 50 - The issue of the cash balance and its pbssible utilization is investigated in the following, where farmers were asked to express their choice for spending such additional income: Preference, percent 1st 2nd 3rd Consumption 50 20 - Construction of new house 20 44 22 Investment in agric. 30 33 56 Investment in other industries - 13 22 Unlike in other survey areas, Mbeya farmers do not seem to have the same high priority for construction of houses, but would .first and foremost satisfy additional consumption. Both investment in agriculture or other industries is increasing from the first to the third preference. In addition to preferences expressed by Mbeya farmers for spending additional farm earnings the following opinions and attitudes are also held: - Generally, farmers are aware of the abaence of fallow land and proper crop rotations, especially when pyrethrum is included. - Secondly, prices received for farm products are considered not adequate. This is especially so in the case of maize, but is also relevant for coffee and pyrethrum. - Thirdly, the limited availability, as well as the high prices paid for fertilizers and pesticides represent a constraint to farming in the eyes of the investigated farmers. Apart from producing fruits and growing potatoes, should the two products find a market, and the keeping of an increased number of livestock, no indication was provided by the Mbeya farmers as to how they would attempt to improve their economic position. 51 - ;,or programming 1beya farms, present activities are used. For maize both the aver1g and best productivity levels are included. Const-27Ants refer to the average farm size of 1.68 ha, the labour availability of 3.6 man-ega.valents (less an assumed 27 percent for livestock keeping) and working ca,p-tal identified as variable costs amountine to 564 shs. First results indicated a cropping pattern of 095 ha. maize, using the best productivity alternative, and 0.73 ha of beans. Both farm land and the level of working capital are completely used up. A gross margin of 6,767 shs is indicated, which is 56 hi'her than at present. Also productivity is up by the same percentage. Wlithout restricting the level of variable costs, the cropping pattern includes 1.49 maize, using best productivity, and 0.19 ha pyrethrum (350 shs variable cost per ha. were assumed for the latter). The gross margin raises to 7,877 shs per farm. 807 shs working capital is required and, besides land, labour during the month of June is constraining. As for the two previous survey areas, the application of practices of best performance in maize cultivation, which consists of higher seed rates, fertilizer and pesticide application, as well as higher labour inputs, and the increased use of working capital on the farm as a whole, farm earnings on present acreages could be sizeably increased. Even without the inclusion of possible returns from livestock keeping, farming returns approximate those of present best farmers, who, however, need much larger acreages than assumed for the present exercise (average). Working capital needs to be increased by 243 shs, for which, once again, small farmer credit could be used, accompanied by an adequate supply of agricultural inputs and extension efforts for the spread of practices presently practised by the best performance farmer in Mbeya. 2.1.4 Cashewnut, Coconut and Rice Farming in the Coastal Area The Coastal Area comprises the divisional census areas 0222, 0224 - 7, 0231 and 0253. Information is available from 45 farmers, whose composite characteristics are summarized in Table 2.16. The coastal farmer is 46 years old and has an educational attainment of 2.6 school- years. His family consists of 6.4 members, who form a work force of 3.8 man-equivalents. Farm sizes, averaging 3.65 ha, ar3 relat'.vely large and var'y from 1.29 to 6.83 ha over the 5 productivity groups. Farmers with highest prcduction seem to be older, with less formal education, larger farm-families and larger work forces. They also farm double the average farm size and have a predominance of perennial crops, such as cashew, coconut and oranges in their cropping pattern. Livestock is only kept in the form of poultry, 8.6 birds per farm on average. Variable costs average 518 shs, with seed contributing 89 percent. Labour inputs amount to 1,782 hrs per farm and tend to decline, on a per hectare basis, with increasing farm size. -Values of production average 5,764 shs per farm, to which perennial crops contribute 76 percent. Rice and maize have higher productivity and contribute to the value of production double the percentage they hold in the cropping pattern. Farms with highest values of production get 15,381 shs from 6.83 ha. This is possibly not only due to increase in farm size but also to increase in productivity. To the sum of gross margins of 5,246 shs, off-farm income of 1,072 shs is added, bringing the total family income to 6,318 shs. The value of home consumption, which is 51 percent of total produce, amounts to 2,394. Household expenditures are rather higher than expected (mainly for clothing, bread and flour), which is perhaps due to the nearness of the capital. Average expenditure is more than average total family income. Only farms with highest farm production show a positive cash balance, which is 27 percent of total family income. For the mass of farmers such a balance does not exist or is hidden behind unrecorded additional earnings of the farm family. - 53 - TEL2E 2.16: Composite Characteristics of Coastal Farms - grouped by level of farm production t .Level of farm production Details Very Very Average . high IHigh -Medium Low low A, No. of nbservationris 999 9 9 45 :Tramt 1e-. labour Parmer's age, years 50.4 47.4 48.4 1 42.9 43.0 46.4 Farm2r's education, school years 1.6 1.8 1.2 5.0 3.6 2.6 Farm faiily, people 9.3 5.7 5.9 7.7 3.6 6.4 Per, hired labour, nos. - - 0.11 - - - 2otal people on farm 9.3 5 7 6.0 7.7 3.6 6.4 Labour availability, man-equiv. 5.3 4.0 3.6 4.3 2.1 3.8 Farm size, ha 6.83 3.86 3.32 2.94 1.29 3.6 (standard error) (1.74) (1.10 (0.44) (0.54) (0(49 Cropping pattern, o Perennial crops 94.3 72.7 72.7 70.3 77.4 79.0 Rice, maize 4.2 21.3 17.7 12.6 16.2 15.3 Cassava, sweet potatoes 1 .5 6.0 9.61 17.1 6.4- 5.7 TOTAL__ 100.0 100.0 100.0 100.0 100.0 100.0 Po,ltry,nmbers 18.0 2.8 4.8 11.4 4.1 8.6 Variable inTuts, shs 1,084 350 466 454 236 518 Perccntage distribution: Seed 96.6 91.1 96.4 79.3 52.2 89.4 Hired labour 3.4 8.9 3.6 18.7 47.8 10.2. Hired power - 2:0 4- 0.4 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 LAbour inputs, hrs 2,690 2,086 1,899 ,492 937 1,782 (Crop production only) Value of farm production, shs 115,381 5,413 4,148 635 1,241 5,764 ( tandard error) ((601.1) (101.96) (190.2) 1(1,037-51 Percentage distribution: Perennial crops 91.0 40.9 62. 61.8 44.4 75.9 Rice, maize 8.9 38.4 2. 2 217 35. 18.7 Cassava, sweet potatoes 0.1 10 .7 9.5 18.4 TOTAL 100.0 0 100.0 100.0 I 100.0 - 54 - TABLE 2.16 (cont'd) Level 'of farm production Detail IVeryi r VerY Average hih High umdiu Loo low Value of farm production, shs 15,381 5,413 4,148 2,635 19241 5,764 Variable costs, shs 1,084 350 466 454 518 Sum of gross margin, she 3 (farm income) 14,297 5,063 3,682 2,181 1,005 5,246 Off-farm income, shs 998 623 2,761 898 80 1,072 Family income, shs 15t295 5,686 6,443 3,079 1,080 6,318 Value of home cnsumption, shs 13,683 3,166 2,518 1,775 829 2,394 Household expenditures, shs . 7,419 12,347 3,660 6,555 3,784 6,753 Cash for investment, replacemen41, savings, consumption 4,193 -9,827 265 -5,251 -3,533 j-2,829 LAND PRODUCTIVITY Value of production, shs/ha 2,252 1,402 1,24l9 896 962 1,579 Variable costs, shs/ha 159 91 14t0 154 183 142 Gross margin, shs/ha 2,093 1,311 1,109 742 779 1,437 Labour input, hrs/ha 394 540 572 507 726 LABOUR PRODUCTIVITY Value added, shs/man-equiv. 2,704 1,274 1,027 527 532 1,395 Value added, shs/labour hr. use 53 2.44 1.95 1-52 1.19 2.91 Ratios Value of output to value of input a. Labour not costed 14.2 15.5 8.9 5.8 5.3 11.1 b. Labour costed 4.1 2.2 1.8 1.4 1.2 2.5 Degree of family labour use, % 33 34 35 22 26 33 Degree of commercialisation, % 76 42 39 33 33 59 Farm income as percent of total income 93 89 56 71 93 83 Cash balance as percent of total income 27 negative 4 negativel -55- Almost half of all farmers (47 percent) have off-farm occupations and earn for an input of 122 days a year, 18.80- shs per day. On average this amounts to 1,072 shs/farm. Farms in the coastal area are quite large The farm size structure is as follows: Average farm Percentage of Percentage of Farm size group,ha size, ha. farms total area Under 1.00 0.83 11.1 2.5 1.00 - 1.99 1.52 17.8 7.4 2.00 2.99 2.40 28.9 19.0 3.00 - 3.99 3.34 20.0 18.3 4.00 - 4.99 4.90 2.2 3.0 5.00 and above 9.08 20.0 49.8 13.65 100.0 100.0 The structural distribution shows that larger farms (over 5 ha.) account for 50 percent of the land available for farming. The cropping -)attern is dominated by perennial crops, as shown below. Crop or crop mix Cultivation size Percentage on ha the total cultivation size Cashew - coconut - oranges 1.91 55.2 Cashew 0.49 14.2 Coconut 0.22 6.4 Other perennials 0.11 3.2 Sub-total 2.73 79.0 Rice, maize 0.56 15.3 Cassava, others 0.17 5.7 Sub-total 0.73 21.0 TOTAL 3.65 100.0 1 In addition to crop farming, 47 percent of the farmers keep livestock in the form of poultry, averaging 19 birds per farm. Cashewnut - coconut - oranges 40 observerations of mixed cultivation of cashewnut-coconut and oranges were recorded. Cultivation sizes are large, 2.23 ha. on average, and unlika most other crops, productivity differences can be observed in spite of little variation in cultivation sizes over the 5 productivity groups. Variable costs relate almost exclusively to the hiring of seasonal labour (102 shs). Labour inputs average 324 hours and are only.slightly higher in the highest productivity group. On average 63 cashew trees, 64 coconut trees and 26 orange trees are grown per hectare. The value of production averages 1,603 shs and the gross margin 1,501 shs. In the group of highest productivity these figures are doubled without clear indication as to what this increase is due to, except that the number of coconut trees/hectare is increased (Table 2.16). 1Although Cobb-Douglas production func! o i yielded reasonable relationships, the cte::1 responsible for difference in prodi .tLvity still could not be totally determined. -t-*c.ties of production were as follows: I.nd 0-55*, cashew trees 0.11, coconut 1 0.21. crange trees -0-05, labour 0.1C The sum amounts to 0.92 and the adjusted /I 4 marginal productivity for lar is high and amounts to 881 shs/ha, while . for cashew nat are 2.80 shs/tree, for coconut 5.25 shs/tree and for labour 0.49 /wr. These m.gnitudec have to be compared with factor costs, which seem favourable c .,?lly for land and coconut trees. Sixto.an observations are available a-out cashewnut cultivation in purestand. A-verge cult*vation sizes are 1.48 ha, with 118 trees and 393 labour hours per hectare, yf.elding a value of production of only 395 shs/ha., Even under best productivity, with mor:e trees/ha and significantly higher labour inputsq the returns amount to only 633 shs/ha, although this is 73 percent higher (Table 2.17). TTE 2.17 Inout/output Coefficients by Level of Productivity - Cashewnut - coconut - oranges, coastal area - Level of productivity Input/output coefficients Average Very Very hih Hih Medium Low low o :. of observations 8 8 8 8 8 40 0ultivation size, ha 2.66 .2.14 1.56 3.04 1.77 2.23 K,mrber of trees.ha. Cashew 76 15 123 0 73 63 Coconut 110 19 104, 47 41 64 Oranges_50 i 1 69 14 23 26. Variable innuts, shs/ha S eed 10 12 --3 Hired labour 102 89 131 61 148 98 TOTAL 112 90 133 61 148 102 Labour input, hrs/ha 1/ 388 376 238 19 285 324 Value of production, shs/ha i 3,368 1,848 1,222 738 472 1,603 Variable costs, shs/ha 112 90 133 61 148 102 Gross margin, shs/ha 1,758 1,089 32 3 Ea256 1han 1,09 an677 24 1,501 V Excluding re-planting and tasks other than w..eed control and harvesting. Coconut; in puresta:d is cultivated 1. K ?arrnrs cn 0.7A ha. Trees/ha amvmt to i5d, iabour inputs to 422 hour and yie"dr ich an &verage of 3,634 kgs or 2,253 shj_.> 2:t productivity ha- a co-mparable fiG-ro f 3,477 shs/ha. (Table 2.18) m'-.. 2,18 n/Tt_Co-ffie ts by L-el of' Prod.uctivxit- - Cshiewnvt cocc. xbs, Ccstul area - Coconut L.I-iut/outpu-.' coefficicIts y ~2,t Low LAverage High Low Avera e 'o. ~f rservetiors 8. 6 4 .Cuiltrat.on eize. ha 3 84 _ 8 0.66 o.81 0 A o. of tree ha 144 102 118 155 170 164 Var`able -nuts, shs/ha Hired l.bour 81 50 46 l 306 203 Lato r; ss hrs/ha 3 ontrol 399 80 151 402 378 393 Harv..esting, others 608 44 27n.a m 1,007 124 393 489 n.a 422 Yield kg/ha 683 218 395 5,559 2,353 3,634 Va ue of production, shs/ha 683 218 395 3,477 1,459 2,253 Variable costs, shs/ha - 81 50 4 3203 o s ma n,s 683 137 345 3,431 1,153 2,050 Rice Rice,grown on an average of 0.67 ha by 23 or half of all farmers, has high labour inputs of 1,117 hrs/ha, expenses for seed of 96 shs and for hired labour and power of 100, shs and 5 shs respectively. Vields are low (6.92 kg/ha), but are compensated by high producer prices. Value of production amounts to 2,688 shs and the gross margin to 2,487 shs/ha. Best productivity is achieved through 2-½- times higher seed rates, and twice the average labonr inputs. In this way yields increase to 1,591 kg and value of production and gross margin to 6,173 shs and 5,710 shs respectively. (Table 2.19). Cobb-Douglass production function analysis did not yield useful resuilts. TABLE 2.19 Inut/output Coefficie",s by Level of Productivity - Rice, C: :stal area - L .vel of productivity Input/output coefficients Very Aov erage high I Low Medium Low low No. of observations 55 4 23 Cultivation size, ha 0.51 i 0.78 0.59 0.53 .1.04 _,0.67 Variable inputs, shs/ha Seed 235 91 68 33 30 96 Hired labour 128 - 191 141 38 100 Hired power - - 27 - - TOTAL 463 91 286 174 68 201 Labour inputs, hrs/ha Land preparation, planting 852 361 397 329 185 382 Weed control 329 280 172 114 123 198 Harvesting 991 586 403 423 442 537 TOTAL 2,172 j1,227 972 866 750 1,117 Yield, kg/ha 1,591 932 712 484 208 692 Value of production, shs/ha 6,173 3,619 2,764 1,880 806 1. 2,688 Variable costs, shs/ha 463 91 286 174 68 201 Gross margin, shs/ha 5,710 3,528 2,478 1,706 738 2?487 Minor Crops/ Crop Mixes Apart from the above perennials and rice, only a few alternative crops are grown. Maize-rice This mixture is grown by 3 farmers on 1.16 ha each. Variable inputs amount to 112 shs, of whico 86 are for hired labour. Labour inputs amount to 874 hrs/ha and value of production to 1,228 shs. Both crops of this mixture seem to perform better in purestand. Maize Maize is also grown by 3 farmers, but on an average of 0.32 ha only. Seed amounts to 79 shs and hired labour costs to 320 shs. Labour inputs are high and average 1,617 hrs/ha. Yields are low (768 kg/ha) but due to higher producer prices than are realised in other survey areas, value of production amounts to 2,763 shs/ha. - 59 - Cassava Nine farmers cultivate this crop on 0.78 ha each. Variable inputs, except fcy 16 she for hired labour do not exist. - Labour inputs average 697 nrs/ha and valie UN production 1,207 shs/ha. Livestock The only form of livestock kept in the coastal area is poultry, Thenty-one farmers or 47 percent keep it in flocks of 19 birds on average, ranging from 2 to '77, Rturn from poultry keeping are not known and are therefore not included in the farm aalys .1, most of the birds are kept in the farm group of highest farm production. Knitry is valued at 13.26 shs a bird on average. Whole farm The value of farm production of coastal farms amounts on average to 51764 hs and gross margins - after deduction of 1,201 she variable costs:- to 5-246 shs. This result is achieved with an average labour input of 1,782 hours and a farm size of 3.65 ha. (Table 2.20). Farming returns range from 1,065 to as much as 34,297 Ws-, the- latter being achieved by 20 percent of all farmers, on farms of 6M83 han on everage= Farm sizes in this group are almost twice the average but real gains accrue alsc from increases in productivity, which is 54 percent higher than average- In Table 2.20 farms are equally grouped according to farm size. It can be seen that the absolutely largest farms are not necessarily the best in comparing level of farm production, which confirms above interpretation. Cobb-Douglas production functi)n yielded elasticities of 0*42* for land, 0.31* for labour, 0.18 for variable inrputs %I 0.13 for poultry. The sum of elasticities amounts to 1.04 and the adjusted R is 0.56, Marginal productivities for all variables are larger than factor costs, except fo.o labour, for which the two are equal. Farm land should preferably be increased, if additional quantities are available. In Table 2.26 the income situation of coastal farms is summarized. Farm incme average 5,246 shs, to which 1,072 she off-farm earnings have to be added. The resulting family income of 6,318 shs is - even on average - not sufficient to cover the level of household expenditures. The latter amount to 6,753 shs, 29 percent being spent on clothing, 21 percent on breed and flour, and a variety of other expenaitures accounting for the rest. ' .60 - siz cd balan:eac,--za- a,y -- a, ÷ -.÷:.h,hs iam ±etF :- > ::-s-.o c shr..-..-t - 2 :e..; ;r . - -= :M . .zcaee o -n orded .d t:oa ening:or vr-st at. ho h xend:-tuires. TIkLE22 a nruticv~.imut PCcoif::iento for av:ns -a svel of Producýtion LýV)rer I armeI oll hii H- d- um1 Low Ou aý` c, ia hý n , qoiN - , ,s 6 C,' 8993 IvLirr nmber 18 0 8 4 1 8-6 Labour - - iig5 cop hr 690 2 08 99 92 79 Valný of p du i.r shs 5 33 '8 6 23_5 241 576 ca: -,.' os-, ':s i 066 26 8 G s m ns406j 682 2 1 1 6,-5,2 Tinde r i400 00 300 00 -00& 9 29 n 1 230 Po a 2 0,-d80 0 i. o s 2 242 0 32 198 53 235 12243 1 20C 2,i 'APT. 2-PT: (a) Farm and Fimily Income and Household Expenditures - Coastal areas (shs) - Level of farm proc.uction Detail VeryI Very Average - h _______ ___ _ high High Medium Lo r low Vale of farE2dction 157381 5,413 4,148 2,635 . 1,241 5,764 va-riable cos's 1,084 350 466 454 236 518 8rm of gross margin ('arm 14,297 5,063 3,682 2,181 1,oo5 5,246 T --4 Off-farm income 998 623 2,761 E98 80 1,072 F-mily income 15,295 5,686 6,443 3,079 1,080 . 6,318 Value of home consumption 3,683 3,166 2,518 1,775 829 2,394 Household expenditures 7,419 12,347 3,660 6,555 3,784 6,753 0-sh for investmnt rIplacement savinas, 4,193 -9,827 265 -5,251 1-3,533 -2,829 (b) Composition of Hou ehold E7,=enditures Item Shs Percentage on total Clothing 1,967 29.1 Bread, flour 1,435 21.2. Sugar 843 1 12.5 Condiments 814 Transport 749 Kerosin 374 5.5 Stimulants 296 4.4 Tea 55 3.6 Tobacco 241 0.8 Detergents 142 2.1 Household goods 133 2.0 6,753 100.0 Farmers' attitudes Farmers in the coastal area are concerned specifically with both the price and market arrangements for cashewnuts. Almost all farmers complain about the service situation concerning the disposal of this crop. The same attitude with regard to other crops, notably oranges and coconuts, is not very widespread. Viuh regard to agr:c-ltural puts, srme 2 eres of ths farmers ar. dieat,sfed with the supply situation of fertV.,1r .wic. sr p ent alsc _c.ensilder the unavaability of tractor zervc:.ces, ;~m -i . 1 : ".>'ts. .s ristr,.nt.s t th-:..r farmi.ng. New enterprises for the noastal areas ar- not .frentiond, for poultry keeping, investment costs are impeding the,r .xpanc::cn. whL e c.ur Js crnjidered by 10 percent of the farmers to be a constreint -c nr s i., t1-at·Cl All constraLnts removed. coastal farmers `,ov' d -dopt the fw..n cropping pa ern: hr a,roent Cashew- coconat 6,0 51-7 Rice 3,92 33 -4 Cassava O-56 4 ,8 Maize 3 3.0 Others 7I Tota' 1~,72 1000 Although the total farm size is mlch ,i ;xcezs f present average7 farms of s size exist in the area. Also the des-red. prtortis of crops are not too diffe-ent from the present; only rice cultivation is giv:n higher prierity. In addition to crops, poultry would be kept by 20 percgnt of all farmers and- in flocks of 65 b-rd, each. Priority in use of addtional farm incomes :o expre,sed by the following preferences: 2n-d 3r Investment in agricultire 67 25 18 Construction of houses 26 40 11 Consumption 4 24 54 Investment in other businesses 3 9 5 Others (savings, marriage) 2 12 First and foremost farmers would use additional funds for investment in farmi.ng. Second on the list is construction of housAing, 2hese items sattsfied, aditional consumption ranks highest. 63 Farm programmin In programing coastal farmn bo1,,-, a ? £,::r sivares 'nri z p vr: used. The imputed activities relate to :r aiz *tD fr j: ob L)-th r-dc Znd cashew nut - coconut. In.addition r b:.'-h pra_h-,ts and coconuts in purestand and cassavý. ''r- ee:: : e ' T... 2ver g2 9e- , C-, of 3.65 ha, the average level of wcrki:sjcaT.. '..:sL.. 7 and t available family labour, calcula93d , i75 -..: >f po bem-. . . Because of their superiority !n l2,: . e- .C r1 i r jv:: v t hec : ' %.'-1 rice and coconuts of best productivity d1j :. T, -9 'r_1C: U :Livt n :,7..es of 1.03 ha and 2.62 ha respeotively. TI:. j v. , 7.. -. 2- i ea s of coastal farmers about desired crop distr: _t.ci T. :n of gvs E "r*:.ns crounts to 15,119 she, which is the level achievzd < V:'re :x.t - Fern. prciti r., on 6.83 ha. Labour is restraining `ri 4p:: . : . : ch:'t vupplv Jrn MXay a fug_st, - In a second attempt, the cultivat. i ' : is åimitd xhe -oresen average of 0.38 ha/farm. Additional labour h-_s bo1.. :ne ::,r abOve months The ll acreage assumed available for rice cultivat:.n 'culi o us, 'o r,adit:con coconzt cultivation should be expanded to 3.27 ha. 1e r i mar,in 1z,id dcr t3 shs, but neither working capital nor labou:: -eo~11 luntrk,i T2h-2 coastal -rea thus seems to be an area where credit would not ne,eses -ily improve farming resuIt at least not until completely new and more ca,ital rquril -' techno-ge for this farming system are developed and propogated. 2.1.5 Maize and Rice Farming in Morgoro The Morogoro area consists of the div.:or-1 ensu a- ms C915-6, 934-ý . Complete production records are available fre. 71 tarroes abe 2,22 smmarizes the-.r composite characteristics. - 64 - The Morogoro farmer is 43 years of age, has 1.6 years of formal schooling and is heading a family of 4.5 people. No permanent hired labourers are observed.. The :.ork force available for agriculture is 2.6 m-.n-eq:4valents with little difference and no specific trend between farm groups strt:-'.s ,, -"eir level of farm production. Differences exist between these groups in their farm sizes. While the average farmer has just one hectare; higher values of production are produced on larger farm sizes. Three fourths of the cultivated area is devoted to cereal production, of which maize grown in purestand holds the largest proportion. Fei, differences in the crop composition exist between farm groups. Better farms seems to have less rice and sorghum but more of the crop mix Maize/Sorghum and 'other' crops than on average. No livestock were recorded in Morogoro. Variable costs, as well as labour inputs, relate to crop farming only. Variable costs average 223 shs/farm and are used in all farm groups for seed and to meet seasonally hired labour charges. Labour inputs amount to 2,267 hours/farm on average and 3,000 hours on the highest productivity farms. The value of farm production averages 1,971 shs and ranges from 388 to 4,633 shs ovef the farm groups. The contributions individual crop enterprises make to the value of farm production are closely related to their proportions in the cropping pattern. The sum of gross margin's or farm income averages 1,748 shs, ranging from 298 to 4,328 shs/farm from the lowest to highest farm group. This increase has little to do with returns to scale, as productivity follows the same pattern: the better and best farm groups not only farm larger acreages but also reach higher production per area of land. Farm incomes are supplemented with off-farm incomes, which averages 784 shs and is uniformly high in all farm groups. Due to this additional income, total family incomes do not have the same divergencies as when farm incomes are considered alone: the range is from 1,028 shs to 4,856 shs/farm and averages 2,532. Deducting 1,321 shs worth of subsistence production and 932 shs of household expenditure, the cash balanee is derived and this averages 279 shs or 11 percent of total income. It has to be realized, however, that more than 70 percent of the average cash balance accrues to the 20 percent best farmers, and that the rest - owing to curtailed consumption and expenditures - operate around the break-even point. U SUMMARY TABLE 2.22 (cont d) Level of farm Droductio n Detail Averags Very i VerY I 9h- J Medium'.] Lo low Value of farm prod., shs 4,633 ;v328 lt477 838 388 1,971 ')3 1 (50.2) (198.2) (standar error) (362.0) (44-8) (48.9) Percentage distribution: Maize 43-0 4M 35.2 4.6.8 31.9 41.2 Rice 18.4 19.9" '27.2 37.9 15.4 21.8 Sorghum 8,1 18-4- 26.8 17.2 14-9 14-5 Maize/Sorghum 13:6 6-5 - 27,0 8.8 Pulses 5 3.1 4-3 Others l :1 .9 9- 63 10.8 8 TOTAL 100.0 100.0 100.0 1.100.0 100.0 100.0 Value of farm prod., shs 4,633 S26 1,477 838 388 1,971 1 Variable costs, shs 305 214. 193 go 223 Sum of gross margin, shs (farm incoLae) 4028 22019 11263 645 298 1,748 Off-farm income, shs 528 965 803 754 730 784 Family income, sha 4,856 2v984 2,o66 11399 11028 2,532 Value of home consumption, shs 21834 1,6oo 11129 617 316 1021 Household expenditures, shs 1,o67 1,429 834 827 624 932 Cash for investment, I replacement savings, consumptio I n 955 --45 103 -45 88 279 LAND PRODUCTIVITY Value of production shs/ha 21693 29097 11717 1,496 625 19990 Variable co;ts, shs ha 177 278 248 344 145 225 Gross margi*n. shs/ha 2,516 1 819 1069 19152 48o 19765 Labour input, hrs/ha lt746 2 717 2P917 2t314 21341 29289 LABOUR PRODUCTIVITY Value added, shs/man-equi 1,644 673 570 334 261 709 Value added, shs/labour hlvr. used 1-48 0-71 0-54 0-57 0.18 0.31 Ratios Value of out-put to value of input a. Labour not costed 15.2 7:5 6 .9 4.3 4.3 8.8 b. Labour costed 1-5 1 0 7 o.6 o.6 0.3 0.8 Degree of family labour use, % 74 63 0 39 37 58 Degree of commercialisation, % 39 31 24 26 1 19 33 Farm income as percent of total income 89 b7 61 46 1 29 69 Cash balance as percent of totell income 19 1 neg. 8 11 - 67 - The average farm size is just one hectare and farm siees are distributed as follows: Average farm Percent of Percent of size, ha farms total land Under 0.50 0.33 30.0 10.4 0.50 - 0.99 0.72 30.0 21.2 1.00 - 1.49 1.15 18.6 21.4 1.50 - 1.99 1.64 10.0 16.5 2.00 - 2.49 2.17 5.7 12.5 2.50 and over 3.12 51 18.0 TOTAL 1.00 100.0 100.0 This distribution reveals disparity in land availability. Sixty percent of all farmers cultivate 32 percent of the total farm land. The cropping pattern is restricted to annual crops only. The main crops or crop mixes observed were: Cultivation Percentage Crop, crop mix size,ha on cultivation size Maize 0.39 39.0 Rice 0.23 23.0 Sorghum 0.14 14.0 Maize/sorghum 0.09 9.0 Cotton 0.06 6.0 Sunflower 0.03 3.0 Minor crops (beans, cassava, sweet potatoes, maize/cassava, groundnuts, 0.6 6.0 tobacco, bananas, coffee) TOTAL 1.00 100.0 - 63 - Eightyfive percent of cultivated area is devoted to staple graina. Their percentage tends to decrease with increasin- farm size. Of the various grains, sorghum seems to prevail on smaller farms. More 'sther' crops can be grown on larger farms where the more basic food needs are readil.- satisfied. tze Fortythree farmers, or 61 percent, gro: maize in pure stand. Cultivation size averages 0.64 ha, ranging from 0.05 to 2.80 ha per farm. Variable inputs amount to 195 shs of which 65 shs are spent on seed, the bulk of 1 2 shs on seasonally hired labour and a small amount of 13 as on charges for tractor ploughing (only 3 fields are tractor ploughed at 218 shs/ha). No other technological inputs, such as, fertilizer, insect/pest control are observed. Thus the only tangible inputs to maize relate to land and labour. On average 2,120 labour hrs/ha are used, 35 percent being used for land preparation and planting, 27 percent for weed control and the rest for harvesting and other tasks. Yield averages 1,791 kg/ha. At producer prices of 1.07 shs/kg, the value of production is 1 916 shs and the gross margin 1,721 shs/hae In Table 2.23 results of maize cultivation are grouped by level of productivity, which is 79 percent higher in the best group than the average. This seems to be achieved through higher seed rates, as all other parameters are similar if not less than average, In order to determine the individual effect of all inputs, production function analysis was carried out. Both land and seed rate have high positive elasticities of 0.49* and 0.25* respectively. The power input has an elasticity of 0.03, labour in land preparation 0.15 and in weed control - 0.14. Their sum amounts to 0.78 with an adjusted R2 of 0.52. Marginal returns are higher than opportunity costs for land (938 shs/ha), seed (8.40 shs*A) and power (4.42 shs/shs). For labour, marginal productivity is small (0.39 shs/hour) and has negative marginal returns when used in weed control. */1 Optimal resource combination requirer >ve all, increases in thom resources which have - compared to their factor costs - the est marginal returns. Dn the case of maize in Morogoro, cultivation sizes shoul' noreased to 1.50 ha; seed rates should be 102 shs/ha, power input should average pre, r levels and labour in land preparation should be decreased. Although overall prc .tion would increase to 2,838 shs, productivity would remain at present levels. TABLE 2.23 Input/output Coeffic -.ts by Level of Productivity - Maize, Morogoro - Input/output coefficients . of productivity Average Very hi Low L low No. of observations 8 899 1 43 Cultivation size, ha 0.66 0. 9 2 0 0.62 0.57 0.64 Variable inputs, shs/ha Seed 71 68 5o 5o 42 57 Fertilizer - - - - Pesticides - - Hired labour 94 118 238 106 31 112 Hired power 15 - 30 29 - 13 Others 1 29 22 1 3 13 TOTAL 182 2.5 340 186 76 195 Labour inputs, hrs/ha Land preparation 720 661 1,040 855 495 736 Weed control 363 588 604 693 622 574 Harvesting, others 639 662 1,114 711 ,C01 797 TOTAL 1,722 1,911 2,758 2.259 2,118 2,107 Yield, kg/ha 3,201 2,20 1,619 1,192 537 1,791 Value of production, shs/ha 3,425 2,360 1,732 1,275 575 1,916 Variable costs, shs/ha 182 215 34c 186 76 195 Gross margin, shs/ha 3,243 2,145 1,392 1,089 499 1,721 Rice This crop is grown by 25 f rmers or 35 percent of all Morogoro farmers. Cultivation sizes average 0.60 ha. Variable inputs average 3f6L shs/ha, of which 84 shs is for seed and 195 shs for seasonally hired labour. More than 1/3 of all rice growers use hired power, for which charges of 231 shs per ha. hav. to be met. - 70 - Labour inputs amount to 2,811 hrs/ha, off which 19 percent is contributed by seasonally hired labour. 949 labour hours or 37 percent of the total input is need in land preparatior and plantingv 487 hours or 17 percent for weed control and the rest for harvesting and other operations. The yield averages 1,723 kg/ha yielding at producer prices of 1.08 shs/kg, a value of production of 1,861 shs/ha and a gross margin of 1,497 shs/ha. Unlike maize, rice yields diminish continuously with increasing cultivation size and a similar pattern is found for the labour input (Table 2.24). Rice farmers using hired power in land preparation have average cultivation sizes and seed rates. Charges for hired power amount to 233 shs/ha. The labour input for land preparation is only 205 hrs/ha or 16 percent of the unmechanized observations. Also weeding decreases to 297 hours or . percent. Yields are down by 5 percent and value of production at the same producer price amounts to 1,787 shs/ha. If for the sake of comparison the different amounts of seasonal hired labour use are omitted, gross margins for mechanized farmers averages 1,476 shs, while the unmechanized farmer.- not burdened by the costs of mechanization - achieves 1,795 shs/ha. The labour requirements for the latter are, however, more than doubled, and the observed increased use of hired labour suggest that a real bottleneck in labour availability seems to exist, which can be broken by using tractor ploughing* Returns to labour amount to 0.88 and 0.49 shs/hr for the mechanized and unmechanized rice cultivation respectively. Table 2.2.4 also suggests that productivity can be doubled on smaller than average cultivation sizes using higher seed-rates and much higher labour inputs than average. Cobb-Douglas production function analysis was performed to determine factors which contribute most to raising farming returns. All variables included in the analysis produced positive elasticities, the highest of which are reached by land (0.46*) and seed rate (0.41*). Power input yields an elasticity of 0.13 and labour, both in land preparation and weed control has 0.04. Thus the sum of elasticities amounts to 1.08 with an adjusted R2 of 0.53. - 71 - Corresponding to the above, marginal products are highest for land (856 shs/ha), seed (9.15 shs/shs) and power3.15 shs/shs). For labour marginal products amount only to 0.07 shs/hr in land preparation and'0.15 shs/hour in weed control. Optimal resource combination - restricted to present total costs - suggest that productivity and production'could be raised by increased cultivation sizes - where possible - up to 2.24 ha, tripling seed rates and slightly increasing power inputs above average. Labour should be diminished to lowest levels possible. This adjustment could lead to a rise in total production to 5,826 sLs or a productivity of 2,601 shs/ha, which is 40 percent higher than the present average. TAtLE 2.24 Input/Output Coefficients by Level of Productivity - Rice, Morogoro - jInput/Output Level of productivit Mechan- Unmeohan Coefficients iVery eized ized high High Medium Low low No. of observ. 5 5 5 5 5 25 8 17 Culf.size,ha 0.35 0.49 0.90 0.5 0.69 0.60 0.61 0.59 Variable inputs, shs/h Seed 131 112 84 51 65 84 78 86 Fertilizer - - - 8- - - Pesticides - -- Hired labour 756 355 82 1151 14 195. 62 257 Hired power 35 63 145 68t 19 76 233 - Others 1 16 2 29 4 9 - 13 TOTAL 923 5461313 263 102 364 373 356 Labour inputs, hrs/ha Land prep, planting 3,757 803 j 325 793 593 949 205 1,299 Weed cont. 1,378 289 285 427 497 487 297 856 Harvesting, others 2,292 1-738 1,105 1,282 1,95 1,375 1 ,1661 1,443 TOTAL 7,427 2,830 1715 2,502 2,185 2,811 1,668 3,628 7,2 1,1 ,2 Yield, kg/ha 3,439 2,226 1,865 1,317 650 1,723 1,656 1,754 Value of prod. shs/ha 3,714 2,404 2,014 l,422 702 1,861 i 1,787 1,894 Variable costs, shs/ha 923 546 313 26, 102 364 373 356 Gross margin, shs/ha 2,791 1,858 1,701 1,159 600 1,497 1,414 1,538 - 72 - Sorghum Sorghum is grown by 20 farmers on average cultivation sizes of 0.43 ha. Seed, hired labour and hired power are the main variable inputs, amounting to 136 shs/ha. Only 3 farmers hired power for land preparation and reached highest productivity levels. Labour inputs average 2,461 hrs/ha. Of this, 749 hours are used in land preparation and 579 hours in weed control. The average labour input level lies between that for maize and rice in this area. Yields amount to 1,369 kg/ha. The highest productivity group of farmers reached 2- this average. Produc,er prices are higher than for rice and maize and average 1.57 shs/kg, thus yielding a value of production of 2,143 shs/ha and a gross margin of 2,007 shs/ha, which is 18 percent higher than for maize and 34 percent higher than for rice (Table 2.25). Highest productivity is reached on smaller acreages than average and with substantial increases in all inputs: higher seed rate, tractor ploughing and neve:btheless higher labour inputs for both land preparation and weed control. -Contrary to maize and rice, the results of a Cobb-Douglas production function for sorghum does not indicate increase in acreage as one of the measures for improving farming returns. Land has a negative elasticity (-0.03). A higher elasticity is obtained for both power input (0,53) and labour in land preparation (0.46). The elasticity for seed is 0.13 and for labour in weed control is 0.28. The sum of all elasticities adds to 1.37, and the adjusted R2 amounts to 0.56. Marginal products are higher for power inputs (37.84 shs/shs) and seed rate (4.64 shs/shs). For labour, marginal productivities are near wage rates and amount to 1.32 shs/hr for labour in land preparation and 1.04 shs/ha for labour in weed control. For power the marginal returns are disproportionately great (compared to their opportunity costs). This makes unrealistic an attempt to calculate optimal resource combination. It seems, however, that one can recommend maximum possible power use along with higher seed rates as possible means for increased production. - 73 - TABLE 2.25 Input/output Coefficients by Level of Productivity - Sorghum, Morogoro - Level of productivity Average i Input/output coefficients Very LrVery high Hi gh Medium Low low No. of observations 4 4 4 4 4 20 Cultivation size, ha 0.33 - 0.29 0.45 0.64 0.44 0.43 Variable inputs, shs/ha Seed 92 91 53 42 49 60 Fertilizer - - - - - Pesticides - - Hired labour 65 66 37 - 34 Hired power 138 68 - - - 30 Others 45 . 12 18 - 12 TOTAL 275 236 137 79 49 136 Labour inputs, hrs/ha Land preparation, planting 1,003 1,206 731 774 238 749 Weed control 989 961 642 368 264 579 Harvesting, others 1,373 2,273 1,538 454 774 1,133 TOTAL 3,365 4,440 2,911 1,596 1,276 2,461 Yield, kg/ha 3,1921 2,172 1,333 740 425 1,365 Value of production, shs/ha 5,01l 3,410 3 2,092 1,161 667 2,143 Variable costs, shs/ha 275 236 137 79 49 136 Gross margin, shs/ha 4,736 3,174 1,955 1,082 518 2,007 Maize/Sorghum A mixture of maize/sorghum is grown by 7 farmers on average cultivation size-of 0.67 ha, which is larger than those for both crops in purestand. Inputs of seed, power and hired labour lie between those for both crops in purestand. The little difference there is, is responsible for larger values of production as well as gross margin. Both crops in the mixture seem to have a complementary effect. Best observations inOicate higher returns to be coupled with higher seed rates and higher labour inputs (Table 2.26). - 74 - TABLE 2.26 Input/Output Coefficients by Level of Productivity - Maize/sorghum, Morogoro - Level of' productivity Input/output coefficients Average High Low No. of observations 3 4 7 Cultivation size, ha 0.63 0.7Q 0.67 Variable inputs, shs/ha Seed 87 42 61 Pesticides - 124 71 Hired labour 42 - 16 TOTAL 129 166 148 Labour inputs, hrs/ha Land preparation, planting 949 337 599 Weed control 754 346 521 Harvesting, others 1,413 512 898 TOTAL 3,116 1,195 2,018 Yield, kg/ha - - - Value of production, shs/ha 4,188 795 2,249 Variable costs, shs/ha 129 162 148 Gross margin, shs/ha 3,859 783 2,101 Cotton Cotton ig grown by 6 farmers on average cultivation size of 0.61 ha. Seed, pesticides, packing material, hired labour and power contribute to the variable costs, totalling 145 shs/ha. Labour inputs are considerably higher than for other crops and amount to 544 hrs/ha for land preparation and planting, 1,110 hrs/ha for weeding, cultivating and pest control and the rest for harvesting and handling of the crop. Total labour input amounts to 2,870 hrs/ha* Only 1 farmer ued tractor ploughing and it might be coincidental that he had the highest yields. Yields averaged 712 kg/hao At the average producer price of 1.24 shs/kg, the value of production is 881 shs/ha and the gross margin 736 shs/ha. In Table 2.27 the average is compared with the highest productivity cases. It seems, that higher expenses for pest control, power input, and higher labour inputs - particularly in weed control - are the main determinants of successful cotton cultivation. 1/* - 75 - TABLE 2.27 Input/Output Coefficients by Level of Productivity - Cotton, Morogoro - Level_ of productivity Input/output coefficients Average High Low No. of observations 3 3 6 Cultivation size, ha 0.63 0.59 0.61 Variable inputs, shs/ha Seed 231 53 38 Pesticides 53 43 48 Hired labour - 48 24 Hired power 39 1 20 Others 24 6 15 TOTAL 139 151 145 Labour inputs, hrs/ha Land preparation, planting 618 470 544 Weed control 1,340 880 1,110 Harvesting, others 1,001 1,431 1,216 TOTAL 2,959 2,781 2,870 Yield, kg/ha 1,011 413 712 Value of production, shs/ha 1,254 508 881 Variable costs, shs/ha 141 149 145 Gross margin, shs/ha 1,113 359 736 Minor crops Besides the aforementioned, a variety of minor crops is cultivated, including cassava, maize/cassava, sunflower, bananas, cashewnuts, coffee, beans, sweet potatoes, groundnuts and tobacco. The input/output coefficients of these crops, which have at least two observations, are summarized in Table 2.28. None of these are used in farm programmingo - 76 - TABLE 2.28 Input/Output Coefficients - Minor crops, Morogoro input/output coefficients Cassava Maize/ Sun- Bananas cassava flower No. of observations .2 2 2 2 Cultivation size, ha, 1 0.88 1.14 0.30 Variable inputs, shs/ha Seed 183 125 24 230 Hired labour - - 52 Hired power - - 158 - Others - - 9 TOTAL . 183 125 243 230 Labour inputs, hrs/ha 'Land preparation, planting 1,049 539 263 378 Weed control 353 786 141 1,419 Harvesting, others 203 601 546 281 TOTAL 1,605 1,926 951 2,078 Yield, kg/ha - - 932 - IValue of production, shs/ha 2,610 1,714 742 2,703 cVariable osts, shs/ha 183 125 243 230 Gross margin, shs/ha 2,4Z7 1,589 499 2,473 Whole farm On one hectare of land and with an input of 223 she worth of material inputs and 2,267 labour hours, the Morogoro farmer obtains an average value of production of 2,267 shs. One third of this is in cash. His gross margin amounts to 1,748 shs. 85 percent of this is derived from cereal cultivation, mainly maize. Table 2.29 summarizes the main .input/output coefficients of Morogoro farms grouped by level of farm production and by farm size. Highest production is achieved by larger than average farms, which have also proportionately higher variable and labour inputs. The increase is, however, not only due to increase in cultivated land, but is also due to a 35 percent increase in productivity, This is not achieved through higher input levplo, but through a more rational use of the inputs coupled with a more optimal cropping pattern. - 77 - The grouping of farm results according..to farm size indicates clearly that increase in cultivation size per se does not necessarily maximize farming returns. The largest farm in Morogoro, having cultivation sizes of 3.12 ha, only reached - at higher inniut levels - the same level of farm production as best farms on 1.72 ha. In other words, the productivity of the latter is 76 percent higher. Cobb-Douglas function produced a disproportionately higher elasticity for land in comparison to labour, implying that more optimal combination of resources would have to be sought in disproportionate increase in farm size. In Table 2.30 the income situation of the Morogor farmer is analysed. Farm income averages 1,748 shs/farm. Best farms have 2 g times more and a large number of farms have substantially less than average (298 shs for the lowest farm group). Off-farm income is spread equally over all farm groups and averages 784 shs, which is 30 percent of total family income which is 2,532 shs. The gap between farm groups - due to equal spread of off-farm income - is narrow and ranges from 1,028 to 4,856. From this the value of farm produced home consumption has to be deducted which is valued at 1,321 shs, and-household expenditures of 932 shs have to be met. The value of farm produced home consumption is highest in the best farm group and declines from there onwards. Household expenditures follow the same trend although the decline is more moderate. The bulk of the latter is spent on clothing and household items. Also of importance are tobacco, kerosine, detergents and sugar. The cash balance left amounts to 279 shs or 11 percent of total income. It has to be realized, however, that more than 70 percent of this cash balance accrues to the best farm group only. Of the other farm groups, even at reduced home consumption and/or expenditures, the cash balance is either meagre or does not exist. Only the above average farms -seem to be viable. - 78 - TABLE 2.29 (a) Input/output Coefficients of Farms - grouped by level of farm production - - whole farm, Morogoro - _______Level of farm production ____ Input/output coefficients Level Vr Average Very I very high .,High Medium Low low No. of observations 15 14 14 14 71 Cultivation size, ha 1.72 1.11 0.86 0.56 0.62 0.99 Labour inputs, ha. 3,004 3,016 2,514 1,296 1,452 2,267 Value of production, shs 4,633 2,328 1,477 838 388 1 1,971 Variable costs, shs 305 309 214 193 90 223 Gross margin, shs 4,328 2,019 1,263 645 298 1,748 (b) Input/output Coefficients of Farms - grouped by farm size - - Whole farm, Morogoro - Farm size group, ha Input/output coefficients Under 0-50- 1.00 -i 1.50-2 0.50 0.95 1.49_A 1.99 2.49 & above No. of observations 22 21 13 7 4 4 Cultivation size, ha 0.33 0.72 1.15 1.64 2.17 3.12 Labour inputs, hrs 1,703 1,692 2,489 2,094 5,125 5,102 Value of farm production,shs 1,178 1,505 1,927 2,936 4,435 4,759 Variable costs, shs 174 164 162 395 443 484 Gross margin, shs 1,002 1,341 1,765 2, 3,992 4,272 196 5_____ 3, - 79 - TABLE 2.30 (a) Farm Family Income, Home Consumption and Household Expenditures Morogoro - Level of farm productioi A Detail .Average Very Very high High Medium Low low Value of farm production, shs 4,633 2,328 1,477 838 388 1,971 Variable costs, shs 305 309 214 193 90 223 Sum of gross margin, shs 4,328 2,019 1,263 645 298 1,748 (farm income) Off-farm income, shs 528 965 803 754 730 784 Family income, shs 4,856 2,984 2,066 1,399 1,028 2,532 Value of home consumption, shs 2,834 1,600 1,129 617 1 316 1,321 Household expenditures, shs 1,067 1,429 834 827 624 932 Cash for investment, replacementi savings, consumption 955 -45 103 -45 88 279 (b) Composition of Household Expenditures Percentage of Item Shs total Clothing 300 32.2 Household goods 182 19.5 Stimulants 103 11.0 Tobacco 90 9.6 Tea 13 1.4 Kerosine 81 8.7 Detergents 73 7.9 Sugar 68 7.3 Condiments 44 4.7 Alcoholic drinks 37 3.9 Flour, bread 29 3.2 1 Others l5 1.6 TOTAL 932 100.0 - 80 - Farmers' attitudes The lack of fallow land is a felt constraint by more than 50 percent of the farmers. Only a few also mentioned the absence of suitable land for growing rice or bananas. Land, due to population pressure, has to be almost continuously cropped and grains cultivated on more than 85 percent of the available farm land, do not allor for much of a rotation to be practised.. With regard to restraints on either disposal of farm products or acquisition of supplies, the following major iS8ues were raised. Unsatisfactory producer prices for maize are mentioned by more than 1/3 of the interviewed farmers. Unavailability, or untimeliness of availability, of tractors is equally considered to be a constraint. Finally, high input prices for both tractor use, fertilizer and pesticides, constitute the third major restraint mentioned, along with the lack of know-how for their most appropriate application. It was also mentioned that the use of these inputs, especially mechanization would lower farming returns. As could be seen from the previous analysis, this could very well be the case, when land and labour inputs are not costed. With regard to possible changes in crop areas according to changing product or input prices, substantial increases are anticipated if product prices were increased but no reaction was indicated with regard to changes in input prices. With more than half the total production being for home consumption, and material inputs being largely confined to (home produced) seeds, an analysis of quantitative response was not attempted. The response to questions,about potential enterprises and restraints to their being undertaken in the present farm organization is analysed in Table 2.31. Livestock keeping, in the form of cattle, goats and chickens, is the enterprise most mentioned as having potential. Restraints mentioned are lack of investment money and disease. The next enterprises considered as having potential were cotton, sesame and sorghum, the major restraints being labour and money. Farmers repeatedly reported that no labour (implying unpaid family labour) is available for cultivation of these crops. 'Money' to a large extent implies working capital to nay for hired labour, machine costs and other inputs. For beans, sunflower, vegetables, rice, groundnuts and oowpeas, 'money' was noted as the first restraint followed by labour. r n2 referencce 1 2 Livestock 27 Invest. ( 51) Disoase ( 26) Graz, lnd ( 18 1.c2r ( 5' mfioney Gotton 26 Labour ( 50) Monej ( 30) Famine ( 11) Disc,so, ( 9 lanC Sesame 25 " ( 44) Seed ( 8) Sorghum 13 ( 38) " ( 23) Birds ( 12) Land, ( 23) famine Beans 12 Money ( 58) Land ( 16) Seed ( 16) Labour ( 10 Sunflower 9 "( 55) Labour ( 33)1Farine ( 12)1 Vegetables fl ' ' 5egetablesP9s" ( 5512) Rice 8 (50)" ( 25) Land ( 25) - Sweet potatoes/cassava 7 Labour ( 28) Land ( 22) Pests ( 28) Sted ( 16) Gra r.ts 6 Money ( 50) Labour ( 34) Seed ( 16) - Cowpea 6 ( 34) Seed ( 34) Pests ( 16) Famine ( 16) Maize 2 Labour (100) -- Sugarcane M Market (100)i - - Bananas i Land (100)1 - lf Figures refer to precent frequency. - 82 - From this one can conclude that: i. farmers would like to stock their farms with livestock; ii. they are aware of 'other minor' crops previously mentioned, which at present are predominantly found on the larger size farm groups; iii. farmers are aware that these new enterprises demand more family labour than is available and funds ('money') would be required for hiring labour and/or substituting it partly by using tractors. Farmers also indicated that on average there could be 3 hectares in excess of present acreage available for expansion of their farms. During an individual check, this was also expressed as preference to re-organizing their farms, once all restraints had been removed. An average farm in the area would then look as follows: Cotton 0.77 ha Maize 0.59 " Rice 0.49 " Sorghum 0.37 " Sesame 0.30 " 'Cassava 0.24 " Beans 0.24 " Vegetables 0.20 " Other (sunflower) cowpeas, sugarcane, coconut, millet, groundnuts, bananas) 0.38 3-57 plus Cattle 2 head Goats 14 " Chickens 23 This could,in part, be wishful thinking, but it also indicates the direction in which farms could be re-organized to give people a more rewarding economic base. - 83 - Finally, the preference of farmers in the spending of additional earnings was investigated, which revealed the following situation. Preference, percent 1st 2nd 3rd House construction 48 16 21 Investment in agriculture 26 37 10 Investment in other industries 14 18 26 Consumption 5 19 30 Savings 7 5 9 Others - 3 TOTAL 100 100 100 Apart from the expressed need for better housing, faramrs in this area are investment orientated, a potential which should be utilized by supporting them accordingly. Farm Programming For programming Morogoro farms, the following enterprises enter the LP matrix as activity alternatives: Maize (average and best productivity) Rice (average and best productivity and mechanized cultivation) Sorghum (average and best productivity) Maize - Sorghum (best productivity) Cotton (best productivity) Restrictions relate to the present average farm size of 0.99 ha, working capital (variable inputs) of 223 shs/farms and a family labour supply of 325 hours/month/farm. The optimum solution for maximizing gross margins under the restraints imposed, indicates the following cropping pattern: Sorghum (best productivity) 0.51 ha Maize (best productivity) 0.10 ha Maize - Sorghum (best productivity) 0.08 ha - 84 - The sum of gross margins for this solution amounts to 3,085 shs; 0.30 ha or 1/3 of presently available farm land would remain uncultivated. Family labour supply is constraining during the months of March, April, May and July. Also the supply of working capital runs short, although not yet constraining. In a second attempt, labour in the critical months has been increased under the assumption that hired labour could be employed. The solution obtained indicates that the cropping pattern should include the following: Maize (best productivity) 0.21 ha Sorghum (best productivity) 0.62 ha Even at sizeably increased labour inputs assumed to be supplied by hired labour, the supply of labour becomes constraining in the months of February and July, while the amounts of variable inputs, other than hired labour, does not appear to be constraining. The sum of gross margins increases to 3,611 shs from which 380 shs for the hiring of seasonally employed labour has to be met. The real gross margin thus amounts to 3,231 shs, which is a little more than that achieved by the preceding LP solution, and what is actually achieved by best farmers at present, although on much larger cultivation sizes. In addition, 344 shs over and above present average levels of variable costs would be required. 2.1.6. Maize and Cotton Farming in Shinyanga The survey area of Shin-7anga covers the divisional census areas 1311, 1334 and 1336. The main crops grown are maize-sorghum, and cotton as a cash crop. Both tractorized and ox ploughing are used. 41 farmers were surveyed. The results are summarized in Summary Table 2.32. The same area was previously surveyed 1/, in 1963, and some of the results are compared. Better farmers in Shinyanga have larger families than the average family of 5.3 people, but a work force equal to the average (2.7 man-equivalents). The average farm size cultivated is 2.66 hectares, although best farmers cultivate more than double this hectarage. The percentage of land in staple grains diminishes from almost a hundred D. Von Rotenhan, 'Cotton Farming in Sukumaland'; Smallholder Farming and Smallholder Development in Tanzania. Ed. H. Ruthenberg, 1968 - 85 - percent to about half when farmers are ranked from lowest to highest performance. Only in the latter group does cotton appear to be of real siCnificance in the cropping pattern (20 percent). Comparison with the earlier survey indicates a slight decrease in cultivated farm size from 3.27 he then to present 2.66. Then cultivated land represented 35 percent of total occupied land. In 1963 cotton occupied half of the cultivated size. The value of production amounts to 4,364 shs per farm and is 80 percent higher than in 1963. Best farmers achieve values of farm production almost three times the above average. The value of production over all farms is made up of the same enterprises in almost identical proportions to their cropping patterns. Only cotton contributes a higher percentage to the value of production than its percentage on the cultivated area. Variable costs amount on average to 612 shs/farm (3 times higher than in 1963) and are also three times higher for the gropp of best farmers. Percentagevise there is no striking difference in the contribution of individual items tothe variable costs. Total labour inputs, which are not costed, amount to 2,131 hours on average, but are- twice this figure for the 40 percent best farmers and much less for the poorest farmers. Average farm income amounts to 3,752 shs, and is about one third that of the best farmers. More than 60 percent of all farmers are far below the average. Off-farm income also accrues in larger amounts to the better farmers, thus widening the gap between them and the poorest. Family incomes averages 4,785 shs, from which subsistence consumption, valued at 2,513 shs, as well as household expenditures of 2,221 shs have to be deducted to get the final figure of 51 shs, which represents the cash available to farmers for investment, replacement, savings or additional consumption. A close examination, however, reveals that only 20 percent of all farmers achieve a positive balance, and 80 percent do not reach the target of viability, the level at wh.ch their farm business is self-sustaining. In Table 2.32 certain efficiency ratios are also calculated, which help assess differences between farm gropps: farm income per man-equivalent amounts to 1,423 shs on average. The ratio of farm income of best farmers to this average is 3:1. The degree of family labour utilisation is 50 percent on average, but 90 percent for the top 40 percent of farmers. The degree of commercialization is 43 percent on average, - 86 - but inter-group differences are striking. Only the top group of farmers can afford to market 55 percent of their farm product (60 percent on average in 1963). Differences in efficiency between farm groups are significant and possible explanations are sought in the following. Differences mainly in crop acreage and farm incomes are al.s- obviQus between the present and the 1963 study. The extent to which these differe-> are due to sampling procedures (only cotton farms in 1963), cannot be determined. Also whether these changes indicated are due to population pressure, price policies or real changes in productivity should be subject to follow-up research. The Shinyanga Farmer The Shinyanga farmer is 51 years of age, older than farmers of other survey areas. He has an educationa standard of 2.3 school years. 5.3 people live in his household forming a potential work force of 2.69 man equivalents. 17 percent of farmers engage in off-farm work and are almost exclusively engaged in it: 320 days/year at 16.30 shs/day, or on average over all farmers 764 shs/year. The farm size managed is 2.66 hectares, all land being held in ownershiplike customary possession. Values attached to land are estimated at 1,376 shs/ha. The farm size structure shows significant inequalities as can be seen below: 54 percent of farmers hold only 17 percent of the land, while 25 percent larger farms (4 hectare and above) accumulate about 60 percent of the land. Farm size group, ha Average farm Percentage of Percentage of size, ha farmers total area Under 0.50 0.47 29.3 5.3 0.50 - 0.99 0.98 9.7 3.6 1.00 - 1.99 1.39 14.6 7.7 2.00 - 2.99 2.59 14.6 14.5 3.00 - 3.99 3.72 7.3 10.4 4.00 and above 6.26 24.5 58.5 2.66 100.0 100.0 - 87 - SUMMARY TABLE 2.32: Composite Characteristics of Shinyanga Farms - grouped by level of production - Details Very Level of farm .production Average Very I-Very high High Medium Low low Number of observations 9 8 8 8 8 41 Farmer, family, labour Farmers' age, years 47 54 49 47 58 51 Farmers' education, school-yr1 3.8 165 1.8 2.8 1 2.1 2.3 Farm family, people 6.8 5.8 6.1 4.0 3.5 5.3 Perm. hired labourers - 0.2 - - - Total people on the farm 6.8 6.0 6.1 4.0 3.5 5.3 Labour availability, man-equi 2.3 3.4 3.2 2.4 2.3 2.7 FARM LAND Farm size ha 6.01 3.43 2.59 0.61 0.64 2.66 (standard error) ( 0.81) (0.52) (0.48) (0.12) (0.10) (0.38) Cropping pattern, o(0 Maize, maize/sorghum 55.1 65.0 68.3 96.7 81.2 59.0 Rice 7.2 13.2 903 - - 9.4 Cassava, sweet potato 6.5 12.8 4.6 3.3 9.4 9.4 Groundnuts, maize/groundnuts 9.5 7.0 10.4 - 8.6 Cotton 20.2 44.6 - 9.12. TOTAL 100.0 100.0 100.0 100.0. 1000 100.0 Livestock, head of cattle 5-4 0.9- - 1.22 Variable inputs, shs 1,875 582 499 73 29 612 Percentage distribution: Seed 48.1 587 59.9 13.7 72.4 51:5 Fertilizer 4.5 - - - - 2*8 Pesticides 0.7 - - - - 05 Hired labour 13.7 18.9 11.8 26.0 10.3 1I.7 Hired power 31.0 20.4 23.4 60.3 17.3 28.3 Others 2.0 2.0 4.9 -- 2.2 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 La'jour input,, hrs (not costed 4,062 4480 1,480 412 437 2,131 Percentage distribution: Crops 89.7 69.2 90.8 100.0 100.0 82.3 Livestock 10.3 30.8 922- 17.7 TOTAL 100.0 100.0 100.0 j 100.0 100.0 100.0 Value of farm production, shs 13,351 5,527 2,356 713 304 4,364 (standard error) (64.17) i(1404.39) (354.64) (57.96) (44.99) (820.97) Percentage distribution: Maize, maize/sorghum 57.6 62.1 70.8 96.5 77.3 66.9 Rice 7.9 10.6 7.2 - - 4.8 Cassava, sweet potato 2.3 6.4 4.0 3,5 8.2 3.7 Groundnuts, maize/groundnuts 3.4 7.8 5.1 - - 4.6 Cotton 23.0 4-5 9.0 14.5 16,4 Livestock 5.8 8,6 3.6.-6- 3.6 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 - 88 - SUMMARY TABLE 2.32 (cont'd) Level of farm production Detail VrVeAverage Vey Very hih High JMedium Low lo Value of farm production, shs 13,351 5,527 2356 713 304 4,364 Variable costs, shs 1,875 582 499 73 29 612 Sum of gross marginL shs- (farm income) 11,476 4,945 1,857 640 275 3,752 Off-farm income, shs .1 1,330 2,312 300 562 662 1-,033 Family income, shs 12,806 17,25 2,157 1,202 937 4,785 Value of home consumption, shs 6,132 4,128 1,561 713 230 2,513 Household expenditures, shs 3,226 3,334 1,815 2,037 967 2,221 Cash for investment, replacement savings, consum>tion 3,448 -205 -1,219 -1,548 -260 51 LAND PRODUCTIVITY Value of production, shs/ha 2,093 1,472 877 1,168 475 1,582 Variable costs, shs/ha 311 169 192 120 45 230 Gross margin, shs/ha 1,782 1,303 685 1,048 430 1,352 Labour input, hrs/ha 606 903 519 675 683 1 659 LABOUR PRODUCTIVITY Value added, shs/man-equiv 4190 1,486 59 275 121 1,123 Value added, shs/labour used hr 2.88 1.13 11.29 1.60 0.64 1.80 IRatios Value of output to value of input Labour not costed 7.1 .7 9.7 10.5 b. Labour costed 2.3 0.9 1.2 1.5 0.6 1.6 Degree of family labuur use, % 91 90 32 11 12 50 Degree of commercialisation, % 550 Farm income as percent of total income 89 68 86 53 29 78 Cash balance as percent of total income. 27 negative 1 - 89 - The Shinyanga Farmer The Shinyanga farmer is 51 years of age; older than farmers of other survey areas. He has an educational standard of 2.3 rohool years. 5.3 people live in his household forming a potential work force of 2.69 man equivalents. 17 percent of farmers engage in off-farm work and are almost exclusively engaged in it: 320 days/year at 16.30 shs/day, or on average over all farmers 764 shs/year. The farm size managed is 2.66 iectares, all land being held in ownership-like customary possession. Values attached to land are estimate' at 1,376 shs/ha. The farm size structure shows significant inequalities as can be seen below: 54 percent of farmers hold only 17 percent of the land, while 25 percent larger farms (4 hectare and above) accumulate about 60 percent of the land. Farm size group, ha Average farm size, Percent of farmers Percent of total ha area Under 0.50 0.47 29.3 5.3 0.50 - 0.99 0.98 9.7 3.6 1.00 - 1.99 1.39 14.6 7.7 2.00 - 2.99 2.59 14.6 14.5 3.00 - 3.99 3.72 7.5 10.4 4.00 and above 6.26 24.5 58.5 2.66 100.0 100.0 The cropping pattern includes a variety of crops, of which maize har absolutely the largest proportion. - 90 - Crop, crpp mixes Cultivation size, Percentage on total ha cultivation size. Maize 0.90 36.9 Maize - sorghum 0.34 13.9 Cotton 0.31 12.7 Rice 0.23 9.4 Sorghum 0.20 8.2 Maize- groundnuts 0.13 5.3 Sweet potatoes 0.13 5.3 Cassava 0.10 4.1 Groundnuts 0.08 3.3 Others (tobacco, orchards) 0.02 0.8 2.44 100.0 The cultivated areas almost correspond to the farm land managed, which excludes fallow and grazing areas. Maize Purestand maize is grown by 80 percent of all farmers, the average cultivation unit (there can be more than one field per farmer) comprising 0.86 hectares. From the records available maize seems to be cultivated for several years in succession, only occasionally interrupted by fallow periods. Yields on average amount to 1,294 kg/ha, selling at an average producer price of 1.14 shs/kg. The value of production averages 1,473 shs/ha. Labour inputs amount to 562 hrs/ha. Variable inputs are limited to seed, hired labour and power input, vhich together expressed in monetary terms, amount to shs 157/ha. These vary with different levels of productivity (Table 2.33). - 91 - TABLE 2.33 Input/Output Coefficients by Level of Productivity - Maize, Shinyanga - Input/oLcoefficients of productivity Input/output e Average Very Very high I High Medium Low l_low No. bfcbservations 7 7 78 7 36 Cultivation size, ha 0.94 0.91 0.90 0.68 0.61 0.86 Variable inputs, shs/ha Seed 29 16 21 17 18 20 f Power 143 122 143 61 45 121 Others 12 71 17 12 26 16 iTOTAL 184 145 171 97 89 157 Labour inputs, hrs/ha 'Land prep., planting 153 136 118 338 371 197 Weed control 237 161 137 238 145 175 Others 299 232 203 1 183 134 190 TOTAL 689 529 458 759 650 562 Yield, kg/ha 2,255 1,302 1,073 867 651 1,294 Value of production, shs/ha 2,571 1,485 1,224 988 743 1,473 Variable costs, she/ha 184 145 171 97 89 157 Gross margin, shs/ha 2,387 1,340 1,043 891 65 1,316 Farmers with highest productivity levels achieve almost twice the average yields. Similarly, seed rates are higher as are costs for power inputs. Labour inputs for land preparation are lover, but those for weed control and other labour (mainly harvesting) are qizeably higher than average. Highest productivity coincides also with slightly larger cultivation sizes than average. From the best performance group to the next, productivity drops sharply to nearly 50 percent, and from then onwards it decreases continously, dropping a further 50 percent between the second highest performing group to the lowest. Although costs for ploughinF are the same no matter ihether oxen or tractors are emp'oyed, differentials exitt with regard to performance (Table 2.34). - 92 - TABLE 2.34 Comparison of Hoe, Ox-Plough and Tractor Perforamance in Maize Cultivation - Shinyanga - Hoe Ox Tractor cultivation ploughing ploughing 7o. of observations 7 19 5 Onltivation size, ha 0.41 0.77 1.66 Labour input, hrs/ha Land preparation 514 110 241 Weed control 2T0 171 180 Harvesting, others 183 79 200 TOTAL 907 360 621 Yield, kg/ha 812 1,275 1,709 High elasticities of production are dtained for land (0.31*), variable inputs (0.31*), power input (0.24) and labour in weed control (0.19), while labour in land preparation - due to the mix of technologies involved - had a negative elasticity of -0.04. The sum 2 of elasticities amounts to 1.01 and the adjusted R to 0.33 only. Marginal products are, for all factors with positive elasticities, larger than likely opportunity costs and these factors should therefore, be proportionally increased. An optimal allocation of resources - at the present total cost level - suggests an increase in cultivation size to 1.22 ha, 158 shs variable inputs (seed), 121 shs for power inputs, which means totally mechanized land preparation and 102 labour hours for weed control. The result is a return of 4,646 shs which is 3! times that achieved at present. Also productivity increases sizeably. Maize-Sorghum The second important enterprise is a mixture of maize and sorghum. It is grow,n by 17 percent of all Shinyanga farmers, on 1.4 fields of 1.17 hectares each. This mixture is grol,n most frequently in succession or after maize and is only occasionally interrupted in the rotation by fallow years. Average productivity amounts to 1,304 shs, which is only slightly less than for maize in purestand. Variable inputs include only seed and costs for ox ploughing and amount to 84 shs/ha. The gross margin averages 1,200 shs/ha. - 93 - Also labour inputs are very similar to those for maize and amount to 551 hrs/ha. Only field sizes are sizeably larger than for maite. Due to the limited number of observations and probably the variation in proportions of the two crops, production function analysis did not produce meaningful results. For those reasons, only the major parameters for the 50 percent high and 50 percent low performance groups are compared, (Table 2.35)- TABLE 2.35 Input/Output Coefficients by Level of Productivity - Maize, Sorghum - - Shinyanga - Input/output coefficients Levelhof 4roductivity Average High Low No. of observations 5 5 10 Cultivation size, ha 1.07 1.47 1.17 Variable inputs, shs/ha Seed 17 12 14 Power, shs/ha 52 140 70 TOTAL 69 152 84 Labour inputs, hrs/ha Land prep. planting 318 138 214 Weeding 150 124 143 Harvessting, others 241 174 195 TOTAL 709 436 551 Yield, kg/ha - - - Value of Droduction, shs/ha 1,751 977 1,304 Variable costs, shs/ha 69 152 84 Gross margin, shs/ha 1,682 815 1,120 These data suggest that - contrary to maize in purestand - higher productivity is accompanied by higher labour inputs and less power inputs. Due to the limited number of cases, these results are not conclusive. Cotton Over 20 percent of all farmers groi cotton. The average cultivation size is 1.41 ha. Half of the fields were oxen ploughed and the other half tractor -loughed. In addition, 2 cases were observed using fertilizers and pesticides. Complete records are available for only 6 farms. Miscellaneous costs relate to transport charges, packing material - 94,- and seasonal hired labour. Yields are 1,450 kg/ha on average and the value of production at a producer price of 1.75 shs/kg amounts to 2,450 shs/ha. After the deduction of 362 shs/ha variable costs, the gross margin amounts to 2,178 shs/ha on average. Labour inputs are ,857 hrs/ha. Differences in input/output relationships between ox ploughed and tractor ploughed cotton fields an.presented below. Table 2.36 Input/Output Coefficients by Level of Productivity - Cotton, Shinyanga - Input/output coefficients Levelof technolo9z Average Ox Tractor ploughed plouahed No. of observations 3 3 6 Cultivation size, ha 175 1.96 1.85 Variable inputs, shs/ha Fertilizer 57 72 61 Hired labour 197 43 116 Hired power 143 143 143 Others 19 55 42 TOTAL 416 313 362 Labour inputs hrs/ha Land preparation 167 376 242 Weed control 345 233 314 Harvesting, others 386 313 301 TOTAL 898 922 857 Yield, kg/ha 1,346 1,721 l1450 Value of production shs/ha 2,357 3,011 2,450 Variable costs, shs/ha 416 313 362 Gross margin, shs/ha' 1,941 2,698 2,088 Differences relate to labour inputs which are, although in total almost equal, higher for land preparation/planting for tractor ploughed fields and higher for weed control on oxen ploughed fields. Yields on tractor ploughed fields, achieved on slightly higher cultivation sizes, are 20 percent higher. -9- Cobb-Douglas function analysis - despite only a few observations - gave results .orth considering. Power input produced an abnormally high positive elasticity of 0.75; land and labour indicate fair magnitudes of production elasticities, na:.iely 0.21 and 0.19 respectively; fertilizer surprisingly produced a negative elasticity indicating that higher than present levels would not significantly contribute to raising production. This needs, however, to "be closely investiEtted witA regard to mode, timing and type of fertiliz6r application. marginal returns to land and power inputs are his-her than factor costs. For labour, only returns of 0.56 shs/hr are indicated. In an attempt to optimise resource allocation, power input is restricted to present levels/ha. The results indicate that a less intensive cultivation on cultivation sizes three times larg-er could significantly raise overall returns, although - due to the above restriction on power input - productivity, expressed as value of production per ha would stay the same. The validity of these results, ho.vver, is limited by the small nu-ber of observations. Rice Rice is cultiyated by 20 percent of the farmers on 1.03 ha on average. Value of production is 2,000 Sbs/ha. Po.!er inputs were not used. Innuts of seed amnount to 33 shs and for hired labour 99 shs/ha. Thus the gross nargin/ha is calculated as 1,368 shs. Labour nn-uts amount to 749 hours on averafe. Because of few observations available, data ue only stratified into 50 percent high and 50 rerccnt low perfcramnce. Differences in productivity relate only to high.:r Labour in:u.s for lanG treparation and :lanting and for veedin-; in the better perf r:ing cases. Other causes cannot be determined. -96- TABLE 2.37 Input/output Coefficients by Level of Productivity - Rice, Shinyanga - :Level1 of productivit Input/output coefficients LoAverage High Low No. of observations 43 7 Cultivation size, ha 1.10 1.28 1.03 Variable inputs,shs/ha Seed 33 32 33 'Hired labour 136 67 99 TOTAL 169 99 132 Labour inputs, hrs/ha Land preparation 520 448 458 Weed control 136 50 97 Harvesting, others 198 241 194 TOTAL 854 739 749 Value of production, shs/ha 2,433 ;,416 2,000 Variable costs, shs/ha 169 99 132 Gross margin, shs/ha 2,264 1,317 1,868 Minor crops A-variety of minor crops are cultivated in Shinyanga area, the input/output coefficients of which are summarized below. Sorhum: 15 percent of all farmers grow sorghum in purestand on 1.37 ha. each. Values of production are 1,672 shs/ha; variable inputs of seed, hired labour and hired power amount to 14, 2, 120 shs per ha. respectively. Thus gross margins amount to 1,436 shs/ha. Labour inputs are 508 hrs/ha. Half of the fillds are oxen ploughed, the other half tractor ploughed. Their parameters are compared below. Ox ploughed Tractor ploughed No. of observations 3 2 Cultivation size, ha 1.28 1.96 Labour inputs, hrs/ha Land preparation, planting 149 153 Weeding 163 168 Harvesting, others 149 233 TOTAL 461 .594 Value of production, shs/ha 1,345 2,193 -97- Except for differences in labour inputs (mainly for harvesting and others) and cultivation sizes, no other differences exist in the available data which would explain the differentials in returns. Maize-groundnuts: This mixture on average cultivation sizes of 0.90 ha is grown by 15 percent of all farmers. Data available are not conclusive. The joint value of production of the mix is indicated to be 1,689 shs/ha and labour inputs 564 hrs/ha. Sweet potatoes: As for the above crop mix, data on the cultivation of sweet potatoes did not produce a clear picture. Grown by 27 percent of all farmers on 0.47 ha each, labour inputs were 790 hrs/ha. Values of production seem under-estimated and amount to 683 shs/ha only. Cassava: Grown also by 27 percent of all farmers on cultivation sizes of 0.35 ha on average, cassava produced a value of production estimated at 1,194 shs/ha. The major input is labour, amounting to 790 hrs/ha on average. Ox ploughing was used by 4 of the 6 cassava cultivators. Comparative data of both hoe and ox cultivation are as follows: Hoe Ox cultivation Ploughing' No. of observitions 6 4 Cultivation size, ha 0.24 0.5 Povesh ,t, shs/ha - 143 Jabqufr Mnput he rsha. lnt.hJan-is;- 1 590 179 Weeding 175 128 Harvesting, others 408 229 TOTAL 1,173 536 k Value of production, shs/ha 1,434 1,208 Gross margin, shs/ha 1,434 1,065 - 98 - Groundnuts Seven percent of all Shinyanga farmers gPow groundnuts on 0,85 hectares each. All fields are either tractor or oxen ploughed. Labour, inputs are 475 hrs/ha and values of production 1,003 shs/ha. Observations were too few for detailed analysis. Livestock Six farmers keep cattle,'58 head in total, valued at shs 316/head. Sixty percent of this total, however, is held by one farmer. It may be for this reason that aattle keeping in the following analysis has no significant contribution in determining farming returns. On average, the value of production of cattle keeping on such farms amounted to 1,764 shs, whereby the above cattle keeper accumulated more than 50 percent. On average over all farms it accounted for 258 shs per farm, or for only some six percent of the total ualue of farm production. In addition 9 farmers kept goats, an average of 11.6 head per farm valued at 52 shs/head; 8 farmers kept sheep averaging 8.5 head and valued at 36 shs/head. Finally, 5 farmers kept poultry, 5 birds per farm, valued at 15 shs/bird. More than 65 percent of all livestock keeping farms kept more than one type of livestock. Livestock inventory changes are not included in the following farm analysis. Whole farm The value of farm production for the Shinyanga farm amounts to 4,364 shs on average, with more than 57 percent in kind, that is, to say as home consumption. After the deduction of 612 shs in variable costs, the farm income (sum of gross margins) amounts to 3,752 shs/per farm. In Table 2.38 farm results are grouped according to level of production and farm size. A large proportion of all farms fall way below the calculated average, while a small number exceeds by far this average. Disparity in the level of production and income is large and much more obvious than in other survey areas. Also, the level of farm production seems to be closely related to farm size, although not in a linear fashion; very low farm production is ovident in farms below 1 hectare; about average farm nroduction is achieved between 1 and 4 hectares, while - 99 - exceptionally high farm production is evident,in farm sizes exceeding 4 hectares. Differences within the 1-4 hectare group seem to be compensated by different levels of intensity in resource use (and technology), which keeps farm production in these farms at about the same level. All variables included in a Cobb-Douglas production function analysis produced positive elasticities: land 0.31*; variable costs 0.38*; labour in crops 0.28; labour in livestock 0.01, and livestock numbers 0.13. Their sum is above unity (1.11) indicating increasing returns to scale. The adjusted R2 is to 0.84. Marginal returns are much in excess of unit cost of land and are greater than factor costs for variable costs and livestock. Labour, especially for livestock, has lower marginal products than its estimated opportunity cost. The optimum combination of resources would seem to require the following: increase in farm size to as much as 7.65 ha, increase in variable costs to 1,219 shs, labour inputs of 943 hrs for crop enterprises and similar inputs as at present for livestock enterprises. This would double farming returns to 8,270 shs/farm. The Shinyanga farm is a case where more extensive cultivation would be indicated for increased farming returns. Although productivity (returns/ha) decreases by 1/3 for the same level of total inputs (costs), total farm returns are almost doubled. This is mainly achieved by increasing the farm size - if at all possible - and the amount of variable inputs (see(, power input) but decreasing the amount of labour employed. Heads of livestock are only slightly increased. In Table 2.39, the income situation of Shinyanga farms is analysed. Farm incone (sum of gross margins) are 3,752 shs per farm on average. As in most other survey areas, divergences in level of farm incomes are large and range from 275 to 11,476 shs between the poorest 2U percent and the highest 20 percent respectively. Off-farm incomes from labouring, alternative employment and business profits amount to 1.033 shs on average4 Better-off-farms also have higher levels 6f this type of income, thus widening the total (family) income gap. Total or family inco-.e amounts, on average, to 4,785 shs, from which household subsistence production and purchased expenditures have to be met. - 100 - TABLE 2.38 (a) Inpu/pjtput Coefficients ior F1rms by Level of Farm Production - Whole farm, Shinyanga -. Input/output coefficients Level of production Average Very Very . High High Medium Low --low No6 of observations 9 8 8 8 8 41 Cultivation size, ha 6.01 3.43 2.59 0.61 0.64 2.66 Head of cattle 5.4 0.9 0 0 0 1.22 Labour, hrs i I Crop production 3,645 3,102 1,344 412 437 1,754 Livestock production 417 1,378 136 - 377 TOTAL 4,062 4,480 1,480 412 437 2,131 Value of production, shs 13,351 5,527 2,356 713 304 l,364 Variable costs, shs1,875 1 582 499 73 29 612 ross_mari, shs/ha 11,476 4,945 185640 275 (b) Input/Output Coefficients for Farms by Farm Size - Whole farm, Shinyanga - Farm size group, ha Input/output coefficients Under 050- 1.00- 2.00- 3.00- 4.00 & 0.cf0 0.99n 1.99 2.99 J 399 above .No. of observations 12 4 6 6 3 10 Cultivation size, ha 0.47 7098 1.39 2,59 3.72 6.26 Head of cattle 0 0 0.67 1.00 4.30 Labour, hrs Crop production 408 292 1,059 1,821 3,171 3,915 Livestock production - - - 1,247 1,547 333 TOTAL 408 292 1,059 3,068 4,718 4,248 Value of production, shs 521 535 3,778 4,404 4,147 10,847 Variable costs, shs 60 88 587 379 948 1,138 Gross margin, shs/ha 461 453 3,191 4,025 3,199 9,609 - 101 - These, especially farm produced consumption, decline with decreasing incomes. Even considering that farms with the least income have almost half the number of people to feed than higher income farms, it seems that they have to curtail consumption and expenditures in order to keep within the limits' of their limited family income. Cash for investment, replacement, savings or increased consumption averages only one percent of total income. From Table 2.39, however, it can be seen that 80 percent of all farmers have negative cash balance or, in other words, they are over-spending. How far this is due to recording errors or reflects the true situation can not be determined. Farmers attitudes Farmers in Shinyanga mention a variety of constraints in agriculture; in general: market, price and transport difficulties are felt by almost all farmers for all agricultural enterprises. Also, insufficient labour availability and variability in rainfall is considered to hamper increased rice production. For sorghum, birds appear to be a problem. Cotton is considered to be insufficiently profitable under current practices. Only for maize are no specific problems felt. Generally, for all enterprises, farmers complain about non-availability of fertilizers (24 percent) and tractor services (53 percent). Removing constraints, farmers indicate a willingness to grow the following crops in the following proportions (results from 61 percent of all farmers): Percent Cotton 30.1 Maize 24.1 Groundnuts 16.6 Rice 15.4 Beans 3.3 Sorghum 3.6 Sisal 2.4 Others (vegetables, tobacco) 1-5 100.0 or 12.8 ha - 102 - TABLE 2.39 (a) Farm and Family Income, Home Consumption and Household Expenditures - Shinyanga - - Shs - Details of income and Level of production Av:ragJ expenditure Very 1Very high High Medium L o w low Gross margin (farm income) 11,476 4,945 1,857 640 275 3,752 Off-farm income 1,330 2,312 300 562 662 1,033 Family income 12,806 7,257 2,157 1,202 937 4,785 Home consumption 6,132 4,128 1,561 713 230 2,513 Housholed expenditures 3,226 3.334 1,815 2,037 967 _2,22 Cash for replacement, investment, savings, consumption 3,448 -205 -1,219 -1,548 -260 51 (b) Composition of Household Expenditures - Item Shs Percentage of total Clothing 474 33.6 Alcoholic drinks 345 15.5 Kerosine 228 13.0 Sugar 240 10.8 Stimulants 201 9.1 Tobacco 180 8.1 Tea 13 0.6 Coffee 8 0.4 Detergents 121 5.4 Condiments 59 2.7 Cooking oil 58 2.6 Household goods 53 2./ Others 109 4.9 TOTAL 2,221 100.0 - 103 - Although intended acreages are much in excess of present ones, the percentage distribution of individual crop enterprises indicates much higher proportions, especially for cotton, groundnuts and r-.ce. Maize and sorghum acreages are smaller. In addition 1 crop enterprises, 51 percent of farmers would wish to keep livestock, in particular cattle, and in much greater numbers than at present. Farmers' preferences in spending additional farm incomes are indicated by the following responses: Preferences percent 1st 2nd 3rd Construction of new house 55.6 22.8 6.9 Investment in agriculture 33.3 62.9 517 Consumption 5-5 11.4 20.7 Investment in other industries 2.8 2.9 10.4 Savings 2.8 - - Others (marriage, etc.) - - 10.3 Farmers would first and foremost utilise additional incomes for the construction of new/better homes and secondly invest in agriculture. Preference for the latter is reinforced by their second and third choice. Consumption increases from the first to the third choice but stays below levels observed in other survey areas* The consistent emphasis of the intention to invest in agriculturo does not only indicate a high awareness of Shinyanga with respect to benefits which could be derived from additional investment, but also demonstrates that Shinyanga farmers are prepared to first invest additional income before increasing their consumption. Farm programming For programming Shinyanga farms, the following activities are included: Maize best productivity) Maize oxen ploughed) Maize tractor ploughed) Maize-Sorghum (best productivity) Cotton oxen ploughed) Cotton tractor ploughed) Rice best productivity), and Livestock - 104 - Restrictions imposed relate to the average size of Shinyanga farms of 2.66 ha, working capital of 612 shs/farm and monthly available labour' hours of 337. The optimum solution included: 1.45 ha Maize (best productivity) 0.46 ha Maize-Sorghum (best productivity), and 0.40 ha tractor ploughed cotton 0.34 ha remain unused, while labour is constraining in the months of February, May and December. The sum of gross margins amounts to 5,330 shs/farm, which is 22 percent more than average at present. In a second attempt labour supply during the constraining months is increased, assuming the availability of hired labour. The cropping pattern then changes to include: 1.65 ha Maize (best productivity) 1.01 ha Tractor ploughed Cotton, and the preseently kept number dcattle. The sum of gross margin increased to 6,820 shs and variable costs become constraining. In fact, through the use of hired labour (41 hours in February, 94 hours in May, and 143 hours in December), variable costs would have already been increased to cover these costs. Removing the constraint on variable costs, the cropping pattern subsequently changes to the following: 2.25 ha Tractor ploughed Cotton 0.40 ha Maize (best productivity), and the presently kept number of livestock. This plan would produce a sum of gross margins of 7,205 shs. - 105 - 2.1.7 Maize, Groundnuts and Tobacco Farming in Tabora The Tabora survey area covers the divisional census areas 1541-2, 1544, 1547. Seventy farmers were interviewed and their responses are summarlzed in Table 2.40. As for the other survey areas, farmers are grouped into 5 equal groups according totheir level of farm production. Tabora farmers average 47.7 years of age but the best farmers are distinctly older. They also have some little edacational experience, larger families, employ some permanent hired labourers, have larger farms, grow tobacco and keep most of the livestock. For the average farmer, his family consists of a total of 5.2 people, forming a potential work force of 3.2 man-equivalents. The (cultivated) farm size averages 1.29 hectares, with 80 percent of all farms clustered around this average Best farms average 2.25 ha. Crops grown aLe predominantly a mixture of maize, groundauts and some cassava. Tobacco is monopolized by the best performing farms. The following inalysis will show that tobacco is the most profitable enterprise and really makes the best, farmers. Variable inputs worth an average of 356 shs comprise - with the exception of the best farmers - about 75 percent seeds, the rest being fertilizers and pesticides. Best farmers spend most on fertilizer and seasonal hired laboiar, mainly in tobacco cultivation. Labour inputs average 1,659 hours per farm. The value of farm production amounts to 2,187 shs for the average farm and is made up for the mass of farmers of the proceeds from the crop mix maize, groundnats and cassava. Only the best and better farms derive their farm production from tobacco and/or livestock keeping. Gross margins (farm income) average 1,831 shs, ranging from 209 to 6,411; however, for four out of the five farm groups, farm incomes are below the average. Off-farm incomes accrue in all farm groups and in about equal proportion and family incomes therefore show a narrower range than farm incomes. Family income averages 2,357 shs. The value of home consumption averages 1,109 shs and the household expenditures 1,198 shs, thus leaving a cash halance of 50 shs, which is 2 percent of total income. As for many of the other survey areas, home consumption declines from high to low farm groups, while purchased household items - with the exception of best farms - remain - 106 - comparable between farm groups. The cash left for investment, replacement, savings or increased consumption on average is small, but positive. It is negativo, however, for the three low farm groups and is only sizeable for bect farms, where it amounts to 39 percent of total incomes. Productivity of land (for crop cultivation.o"-' is relatively low for all but the best farmers, which achieve almost twice the average values/ha. Labour productiv'ty sharply declines fro'm the highest to the lowest performance farms. Family labovr use is generally low and averages 32 percent. The degree of commercialization is low for all farm groups, averaging 49 percent, and is almost nil for the lower farm groups. Fourteen percent of farmers are engaged in off-farm occupation, working an average of 231 days per year at 16 shs/day. Average earnings frr all farms amount to 526 shs. The farm size structure in Tabora shows a rather equal distribution. Farms are neither extremely large nor small and are mainly clustered around the mean of 1.29 cultivated hectares. Farm size group, ha Average farm size Percent of farm Percent of total ha area Under 0.50 0.41 11.4 3.6 p.50 - 0.99 0.75 30.0 17.4 1.00 - 1.49 1.28 37.1 36.7 1.50 - 1.99 1.67 10.0 12.9 2.00 - 2.49 2.38 4,3 7.8 2.49 and over 3.92 72 21.6 1.29 100.0 100.0 - 107 - SUMUaRY TABLE 2.40 Composite Characteristics of Tabora Farms - grouped by level of production - Details L ee Level of farm production Average Very ,Very 1 high High Medium Low low No. of observations 14 14 14 14 14 70 Farmer, family, labour Farmer's age, years 53.3 48.3 45.1 45.5 46.2 47.7 Farmer's education, school-yrs 0.3 05 .- -1 Farm family, people 7.1 5.0 4.7 1 4.0 5.2 Per. hired labour, nos. 0.3 0.1 - 0.1 0.1 Total people on farm 7.4 5.1 4.7 5:1 4.1 5.3 Labour availability, man-equiv. 4.4 3.1 2.7 3.7 2.3 3.2 FARM LAND Farm size, ha 2.25 i 1.44 1.10 1.05 0.64 1.29 (standard error) (0.52) (0.17) (6.10) (0.09) (0.08) (0.13) Cropping pattern, %5 Maize groundnuts, cassava 35.1 6o.8 67.0 50.0 72.9 Maize/rice/millet 33.0 11.9 19.8 10.2 8.5 19.2 Cassava, sweetpotatoes 19.8 23.8 13.2 15.6 18.6 16.8 Tobacco 18.1 - - - - Others - 1.5- 2.2 - ;5.b TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 Lives-tock, head of cattle 24.0 6.9 0.4 - - 6.2 Variable inputs, shs 1,239 205 129 98 108 356 Percentage distribution: Seed 12.7 81.3 70.4 75.1 71.3 28.6 Fertilizer 39.1 13.3 25.0 18.6 10.4 33.6 Pesticides 2.4 5.4 4.6 6.3 5.4 3.1 Hired labour 40.4 - - - 12.9 30.7 Hired power - - - Others 5.4 - - - - 4.1 TOTAL 100.0 10100.0 100.0 100.0 100.0 Labour inputs, hrs 3,095 2,501 1,354 743 602 1,659 Percentage distribution: Crop production 50.0 53.0 84.0 100.0 100.0 65.0 Livestock production 50.0 47.0 16.0 - - 35.0 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 Value of farm production, shs 7,650 1,650 791 525 317 2,187 (standard error) (1784.12)1 (194.02) ( 26.08) ( 16.00)( 16.84) t(1182.25) Percentage distribution: Maize,groundnuts,cassava 5.8 45.0 65.4 79.1 83,5 21.3 .Maize/rice/millet 7.5 5.1 11.7 10.5 8.7 7.5 Cassava, sweet potatoe 2.1 7.4 17.2 10.4 5.8 4.4 Tobacco 41.8 ----29.6 Other crops - 9.9 - - -15 Livestock 42.8 32.6 5.7 - 2.0 35.7 TOTAL 100.0 100.0 100.0 100.0 100.0 -108- SUMMARY TABLE 2.40 (cont'd) e Level of farm production Detail Very Ve Average Very Vry high .._High Medium Low low I Value of farm production, shs 7,650 1,650 791 525 317 2,187 Variable costs, shs 1,239 205 129 98 108 356 Sum of gross margin, shs farm income) 6,411 1,445 662 427 209 1,831 Off-farm income, shs 1,37- 950 714 204 886 526 Family income, shs 7,785 2,395 1,376 631 1,095 2,357 Value of home consumption, shs 2,873 1,169 685 502 313 1,109 Household expenditures, shs 1,874 1,097 1,396 1,013 983 1,198 Cash for investment, replacement, savings, consumption 3,038 129 -705 -884 -201 50 LAND PRODUCTIVITY (crops only) Value of production, shs/ha 1,944 j 771 678 500 495 1,091 Variable costs, shs/ha 551 142 117 93 168 276 Gross margin, shs/ha 1,393 629 561 407 327 815 Labour input, hrs/ha 637 920 1,034 707 940 836 LAND FRODUCTIVITY Value added, she/man-equiv. 1,571 466 245 115 94 606 Value added, shs/labour hr. used 2.23 0.58 0.49 0.57 0.36 1.17 Ratios Value cf output to value of input a. Labour not costed 6.20 8.05 6.13 5.35 2.93 6.14 b. Labour costed 1,76 0.61 0.53 0.62 0.45 1.08 Degree of family labour use, % 39 54 33 13 17 32 Degree of commercialisation, % 63 29 2 __ __ __ __ __ __ _63_ _ 24 49 Farm income as percent of total income 82 60 48 67 19 77 Cash balance as percent of total income 39 5 negative - 109 - The cropping pattern includes a variety of crops, of which a crop mix of maize, groundnts and cassava predominates, as indicated below: Crop/crop mix Average cultivation Percentage of size, ha total Maize, groundnuts, cassava 0.64 49.6 Maize 0.18 14.0 Rice 0.04 3.1 Maize, Millet 0.02 1.6 Maize, Cassava 0.11 8.5 Cassava, Sweet potatoes 0.12 9.3 !'Tobacco 0.07 5.4 Others (groundnuts, sugarcane, orchards) 0.11 8.5 TOTAL 1.29 100.0 Maize - groundnuts - cassava .62 farmers (almost 90 percent) grow this crop mix on an average area of 0.69 ha. On nearly all fields this crop mix is grown in successive rotation. The joint value of production of all crop components amounts to 730 shs on average. Variation from this average is relatively little. Variable inputs amount to 134 shs, of which 82 she are spent on seed, 24 shs on fertilizer, 7 on pesticides, 4 shs on hired labour and 6 shs on power charges. Only two fields were tractor ploughed (for 122 shs/ha), and the yield differences was negligible. Labour inputs average 699 hrs/ha, of which 211 are for land preparation and planting, 144 for weed control and 220 for harvesting. In the following, data on this crop mix is divided into 5 groups accoraing to the level of productivity and whether the crop mix is fertilized or not (Table 2.41). Values of production of fertilized crop mixes are 35 percent higher than of unfertilized ones. Gross margins, due to higher variable costs, are only 22 percent higher. In addition, labour inputs increase from 657 to 765 hours/ha. Returns to the additional labour are 1.30 shs/hour, which is higher than the assumed opportunity cost for this input. Comparison of this result with the achievement of the highest productivity - 110 - level, however, suggests that there must be other factors than fertilizer alone which influence results. On smaller acreages, with substantially higher seed rates, slightly more fertilizer and pesticides than average, and higher labour inputs, values of production are more than doubled. Obviously, the right combination of inputs is necessary to enhancd farming results. This issue is pursued by Cobb-Douglas production function analysis. Negative elasticities for seed (-0.05), fertilizer (-o.o0) and labour in weed control (-0.32) suggest that additional application of those inputs would not increase yields. Land has a high positive elasticity of 0.42* and about equal elasticities are obtained for pesticides (0.43) and labour in land preparation (0.40). The sum of elasticities amounts to 0.88, at an adjusted R2 of 0.50. Marginal productivities are especially high for land, pesticides and labour in land preparation. It is the adjustment in the level of these inputs which will increase yields. The extremely high marginal return to pesticides (45 shs/shs) suggests- that proper pest control, above all, could significantly increase farming returns from the crop mix of maize, groundnuts and cassava. Maize 25 percent of all Tabora farmers - among them 3 percent with more than 1 field - grow maizein purestand. Average field sizes are 0.69 ha. Yields average 1,123 kg/ha, amounting at producer prices of 0.69 shs/kg to a value of production of 773 shs. Variable inputs are relatively high and amount to 275 shs, thus leaving only a small gross margin of 498 shs/ha, (Table 2.42). Labour inputs amount to 644 hours on average, of which 249 hours are used in land preparation and planting, 141 in weed control and 126 in harvesting. In Table 2.42 high and low productivity fields and fertilized and unfertilized fields are compared. A higher rate of fertilizer input than average - other factors kept consistent - does result in higher yields and thus higher returns. As the case of the high productivity groups indicates, however, higher fertilizer inputs coupled with higher inputs for seed, pesticides and labour, surpasses by far the level of yiele realised by fertilizer input alone. Production function analysis produced high elasticities for all material inputs: seed (0.29), fertilizers (0,28), and pesticides (0.12), and labour in land preparation (1.31). Both land and labour in weed control have negative elasticities (-0.29 and -0.71 res-ectively) and have thus surpassed their optimm= levels* More intensive - 111 - cultivation of this crop seems to be necessary. The sum of elasticities is 1.00 and the adjusted R2 is 0.90. A more optimal combination of resources is estimated for those inputs having positive marginal returns.- Using less fertilizer but more of all other resources at the present total cost level could profitably raise the value of production. Better pest control, as well as higher seed rates and more labour for land preparation, are the most important factors in an effort to increase farming returns. TABLE 2.42 Input/output Coefficients by Level of Productivity - Maize, Tabora - Level of productivity) Averagei Fertil- Unfertil- Input/output coefficients ised ised . High Low No. of observations 9 9 18 12 6 Cultivation size, ha 0.72 0.66 0.69 0.90 i 0.27 Variable inputs, shs/ha Seed 60 30 45 38 59 Fertilizer 155 79 117 176 - Pesticides 22 4 13 18 3 Hired labour Hired power Others TOTAL 237 113 275 232 62 Labour inputs, hrs/ha Land preparation, planting 292 206 249 1 241 266 Weed control 161 121 141 1 142 139 Harvesting, others 299 209 254 231 299 TOTAL 752 536 644 614 704 Yield, kg/ha 1,752 494 19123 1,376 616 Value of production, shs/ha 1,209 337 773 919 425 Variable costs, shs/ha 237 113 275 232 62 Gross marin shs/ha 972 224 498 717 363 Maize/Cassava This crop mix_'s grown by 16 farmers on fields averaging 0.46 ha. Values of production average 645 shs, with relatively high variation. Variable inputs amount to 152 shs on average, of which 119 she is used for seed, 26 for fertilizer and 7 shs for pesticides. Gross margin is thus calculated at 493 shs/ha. In the following, results of this crop mix are stratified into high and low productivity, as well as whether or not this crop mix was fertilized. (Table 2.43). - 112 Fertilized, the crop mix yields double the value of production as when unfertilized. But this is achieved with a triplirg of variable expcnses and 50 percent more labour input. The high productivity group, on the other hand, through a more prudent combination of inputs, especiali'- : r almost reachas the output of the fertilized crop mix, and with lowcr varin7a cwti 7--n ses the latter in the magnitude of gross margin. Production function analysis was attempted for this crop mix, but as for the previous crop mixes results are difficult to quantify. Landl fertilizer, pesticides and labour for weeding come out with positive production elasticities (0.39, 0.05, 0.70 and 1.17 respectively). Once again, pest control produces such an unreasonably high ratio of marginal return to opportunity cost that a meaningful re-combination of production resources is not possible. TABLE 2.43 Input/Output Coefficients by Level of Productivity - Maize, Cassava - Tabora - Leve of Fertil- Unfert- Input/output coefficients Droductivity iAverage ised ilised High Low No. of observations 8 7 153 12 Cultivation size, ha 0.42 064 0.46 0.28 0.51 Variable inputs, shs/ha Seed 174 56 119 124 118 Fertilizer 60 - 26 183 - Pesticides 12 2 7 24 3 Hired labour Hired power Others TOTAL -- 246 58 152 331 121 Labour inputs, hrs/ha Land preparation, planting 310 219 252 394 216 Weed control 171 185 180 167 183 Harvesting, others 702 146 015. 69 348 TOTAL 1,186 650 847 1,24o 847 Yield, kg/ha---- Value of production, shs/ha 1,089 388 645 1,140 521 Mariable costs, shs/ha 246 58 152 331 121 Gross margin, shs/ha 843 330 493 809 400 - 113 - Cassava 18 sample farmers grew cassava on 0.42 hectares each. Full records are available from only 7 of them. Because little information was gathered on this crop, no analysis :s possible. Sweetp Dotat o es On average cultivation sizes of 0.10 ha, 10 farmers grew sweet potatoes. Vdlue of production amounts to 874 shs on average, from which the value of home produced cuttings should be deducted (75 shs) to arrive at the average gross margin of 976 shs/ha. Labour inputs average 518 hrs, of which 346 hrs are spent on land preparation and planting and 139 hrs on weeding. Harvesting labour is under-recorded (Table 2.44). The case of sweet potatoes is interesting because it can be show-m that through more labour intensive cultivation in both land preparation, planting and weeding returns can be doubled on smaller acreages. TA9LE 2. InDut/Output Coefficients by Level of Productivity - Sweet potatoes, Tabora - Level of productivity Input/Output coefficients Average High Low No. of observations 4 3 10 Cultivation size, ha 0.06 0.14 0.10 Variable inputs, shs/ha Seed 78* _78* 78 TOTAL 78 78 78 Labour inputs, hrs/ha Land preparation, planting 496 246 346 Weed control 176 126 139 Harvesting, others 100 290 33 TOTAL 772 662 518 I Yield, kg/ha Value of production, shs/ha 1,580 662 874 Variable costs, shs/ha 78 78 78 Gross margin, shs/ha 1,502 584 796 * Assumes equal to average - 114 - Tobacco Eight Tabora sample farmers grew tobacco', which is the only real cash crop in the area. Cultivation sizes average 0.60 ha. Values of production are as high as 9,662 shs/ha on average. Also variable in-nuts are high and amount to 1,852 shs, mainly for seed (92 shs), fertilizer (873 shs), pesticides (25 shs), hired labour (740 shs). In Table 2.45 results are grouped according to high and low productivity. It can be seen that increases in all material inputs as well as labour inputs on comparable cultivation sizes can almost double returns. The extent to which individual inputs contribute to this can be determined by production functinn analysis. TABLE-2.45 Input/Output Coefficients by Level of Productivity - Tobacco, Tabora - Level of productivity Input/output coefficients i Average High ;Low No. of observations 4 4 8 Cultivation size, ha0.62 0.59 0.60 Variable inputs, shs/ha Seed 102 81 92 Fertilizer 1,200 515 873 Pesticides 34 15 25 Hired labour 1,297 134 740 Hired power Others 152 62 _ _122 TOTAL 2,79-4807 1,852 Labour inputs, hrs/ha Land preparation, planting 721 358 588 Weed control 362 218 303 Harvesting, others 1,117 678 852_ TOTAL 2,200 1,254 142 Yield, kg/ha Value of production, shs/ha 12,494 6,648 9,662 Variable costs, shs/ha 2,794 807 1 1,852 Gross margin, shs/ha 9,700 5,841 7,810 - 115 - Cobb-Douglas function analysis (employed despite the small number of observations) showed that all inputs with the exception of pesticides and labour for weeding have positive production elasticities: (land 0.47, seed 0.04, fertilizer 0.08, and labour in land preparation 1.58). Pesticides and labour in weed control have negative elasticities. The sum of elasticities amounts to 1.06; the adjusted R2 is 0.99. Contrary to the results obtained by stratification according to level of productivity, production function analysis and subsequent optimisation of resource inputs indicate in total a much less intensive mode of cultivation. Cultivation sizes should be increased to 1.64 ha, while the level of labour inputs for land preparation per hectare should be maintained. Other inputs seem to be of minor importance. Through such an adjustment total production could be increased to 27,386 shs, while productivity could be raised by some 70 percent. Although these results can only be regardad as indicative, they nevertheless show the great potential in the cultivation of this crol which at present is being cultivated by only a small percantage of farmers. Minor crops -In Tabora, some other crops or crop mixes were observed, and results are summarized in Table 2.46. None of these crops is of much importance. The number of observations on each arc too few to permit meaningful analysis. TABLE 2.46 Input/Output coefficients for Various Minor Crops* Groundnuts Maize/ Details Rice Sugarcane Bananas I Beans Mill t_ No. of observations 3 2 22 2 Cultivation size, ha 0.93 0.18 0.62 0.15 0.58 Variable inputs, shs/ha Seed 29 - 65 200 77 Fertilizer -6- -_67 - TOTAL 29 a.a 65 267 77 Land )rep., plantin& 415 341 156 373 141 Weed control 106 106 103 1 280 1 77 Harvesting 86 n.a 137 noa 151 Others 11A n.a 45 270 176 TOTAL 721 447 441 923 545 Value of production shs/ha 205 1,944 302 4,033 4.74 Variable costs, shs/ha 29 - 65 267 77 Gross margin, shs/ha 176 1,944 238 3,817 397 * Crops with only one observation have not been included. - 116 - Livestock Twelve farmers, or 17 percent of all Tabora farmers, are, in addition to crop farming, cattle keepers. They keep quite large herds, on average 37 head, ranging from 5 to 95 head. Eight percent of the cattle kept are bulls, valued at 420 shs each, 67 percent are cows valued at 339 shs/head and the rest are calves valued at 76 shs/head. Labour inputs average 3,456 hours per cattle keeping farm; of these 2,996 hours are for herding (not dependent on the herd size) and 460 hrs for milking, varying from 182 to 1,460 hours per farm. The only cash input relates to the purchase of minerals, amounting to 62 shs on average. . The value of output from cattle keeping amounts to 4,361 shs per herd, of which 1,328 shs relates to the value of livestock products and the rest to the value of inventory changes. Cobb-Douglas production function was attempted but produced a very low sum of production elasticities. The number of animals had an elasticity of -o.45 suggesting that increased returns are not necessary dependent on increases in number of cattle. The elasticity of variable inputs is 0.39. This magnitude along with the low costs for this input suggest that additional inputs are very much worth the money. The elasticity with respect to labour is 0.48, and is most closely related to the portion of labour spent for milking. The R2 was only 0.24. Seven farmers in Tabora keep goats and/or sheep. Four of them are among cattle keepers. The number of animals kept averages 4.1 per farm, valued at 68 shs per head. Whole farm The value affarm production for the Tabora farm averages 2,187 shs. This amount is made up of 1/3 tobacco, 1/3 livestock and 1/3 other crops. If one considers the fact that tobacco is exclusively cultivated by the top 20 percent of farmers and, in addition, that almost all cattle is in the hands of only 12 farmers (in the two highest farm groups), the great disparity in the absolute levels of farm production and income between the mass of farmers and the few keeping large cattle herds and/or growing tobacco becomes obvious. One thing,however, deserves mention. The best farmers are not necessarily those holding disporportionately larger cultivated farm sizes than average. As mentioned before farm sizes are very much clustered around the mean. Follow-up on those results therefore should be directed to discovering possible restraints which prevent the average farmer from tobacco cultivation and means to provide him with livestock for converting his farm to a mixed one. - 117 -o 1he farm incomo3 of 1,831 shs an avera,3 off-iar.a incom2 of 526 shs is added to give a family Jncome of 2,357 shs. Sin.co off-faria incomes'are sproad over the whole rango of farm groups, the gap in farily incmC is narrowed, although best farmers still hold three timc's the average. The value of home consumption steadily declines with decline in level of farm -roduction. It averages 1,109 shs/farm. Housdhcld expenditures - with the exception of best farms - are relatively constant over all farm groups and average 1,198 shs. Thus the cash balance for the farm amounts to 50 shs. Stratification of farmers according to level of production indicates also that for the mass of farmers the cash balance is negative and that only the best farmers achieve a cash surplus equal to 39 percont of their total income (Table 2.48). Farmers' attitudes In Tabora drea farmers have a common complaint: low prices and insufficient market availability of maize, groundnuts, rice and cassava. For tobacco no such complaints were foice". With regard to production means, 90 percent of all farmers expressed their opinion about fertilizers and seed: 93 percent regard fertilizer prices as too high while the rest say that fertilizer is not available. With regard to seed, 80 percent complain about its unavailability, 5 percent about its delayed availability and the rest about its price. Regarding potential enterprises for Tabora, no new ones were named. For the existing enterprises, opinions were as follows: Rice - rainfall is insufficient and suitable land limited. Millet - demand is low, prices are lov and there is a bird problem. Tobacco - for 84 percent of all farmors, labour is seriously constraining cultivation. Cassava - prices are too low wild pigs are a problem. Cattle T investment costs high. a0 - 118 - TABLE 2.47 (a) Input/Output Coefficients for Farms by Level of Farm Production - Whole farm, Tabora - Level of productivity Input/output coefficients - Average Very Very I high High oMedium Low low No. of cases 14 14 14 14 -14 14 Cultivation size, ha 2.25 1.44 1.10 1.05 .64 Head of cattle, no. 24.0 6.9 0.4 6.2 Labour, total hrs 3,095 2,501 16354 7 3602 1,659 Value of production, shs 7,650 1,650 791 525 317 2,187 Variable costs, shs 1,239 205 129 98 108 356 Sum of gross margin, shs 6,411 1,445 662 427 209 1,831 1 (b) Input/Output Coffficients for Farm by Farm size - Whole farm, Tabora - Farm size group, ha Input/output coefficients Uhder 1.0 -T 2.00 - 3.00 - 4.00 - 500 1.00 1.99 2-99 3.99 4.99 'above No. of observations 8 21 26 7 3 5 Cultivation size, ha 0.4 0.74 1.28 1.67 2.38- i___3.92 Head of cattle - 2.7 4.6 i 26.7 2.7 12.8 Labour, hrs 384 955 1,737 2,997 2,547 3,847 Value of production, shs 411 903 1,679 4,755 4,003 8,377 Variable costs, shs 67 116 263 331 581 2,211 Gross margin, shs/ha 344 787 1,416 4,424 3,422 6,166 - 119 - TABLE 2.48 (a) Farm Family Income, Home Consumption and Expenditure - Tabora, shs - Level of production Details of income and Average expenditure Very Very high High Medium Low low Gross margin (farm income) 6,411 1,24 662 209 1,831 Off-farm income 1,374 950 714 204 886 526 Family income 7,785 2,195 1,376 621 1,085 2,357 lHome consumption 2,873 1,169 685 502 i 313 1,109 Household expenditures 1,874 1,097 1,396 1 1,013 983 1,198 Cash for investment, replacement, savings and 3,038 -71 -705 -891' -211 -50 consumption (b) Composition of Household Expenses Item Shs Percentage of total Clothing 410 - 34.2 Sugar 199 16.6 Condiments 146 12.2 Detergents 142 11.9 ;Kerosin 114 9*5 Alcoholic drinks 86 7.2 Stimulants 82 6.8 Coffee 8 0.7 Tea 22 1.8 Tobacco 52 43 Othere 19 1.6 TOTAL 1,198 100.0 - 120 - All restraints removed, farmers would re-organise their farm businesses as follows: Hectares Percent Maize 1.44 41.9 Groundnuts 0.74 21.5 Tobacco 0.88 25.6 Rice 0.28 8.- TOTAL 3.44 100.0 Cattle 39 head Apart from being ambitious with regard to the absolute size of the farm business, the proportions of the individual crops are not too different from the present practice. However, more farmers (78 percent of the total) would engage in tobacco cultivation and all farmers would keep cattle in proportions as presently held by the 17 percent cattle keepers. Finally, the farmers indicated the use to which they would put additional incomes, This is recorded in order of preference. Preference % lst 2nd 3rd Investment in agriculture 48 77 60 Build house 44 14 6 Consumption 2 - 16 Investment in other industries 1 1 3 Others (marriage, education, debt repayment) 8 13 100 100 100 Tabora farmers are conscious of improving their agricultural resource base by investing in it, should additional income be available. Second to this is the construction of new/better homes. Increased consumption and other usages become important only after investment in agriculture and the need for better homes are satisfied. - 121 - 2arr programming The following activities enter the farm programming exercise of Tabora farms: Maize - groundnuts best productivity and fertilized Maize n n a f Maize - cassava n t n f Tobacco best and average productivity Sweet potatoes best productivity) Livestock estrictions relate to farm land, which is taken at its present cultivated size of 1.29 ha, working capital (variable costs of 356 shs), family labour supply of 400 hrs/month, present size of cattle herds (6.2 animals) and the present areas under tobacco and swoot potatoes, which are 0.07 ha and 0.02 ha respectively. Linear programming maximized the sum of gross margins for the following cropping pattern: Maize - groundnuts 0.45ha Maize 0.49 ha Tobacco 0.07 ha Sweet potatoes 0.02 ha Cattle 3.9 head All crops are picked up in their best productivity levels and not in their fertilised average, which seems to demonstrate the advantage of using packages rather than single inputs, such as fertilizers. The sum of gross margin averages 2,191 shs per farm, which is a little over what is realised by the average Tabora farm at present. 0.27 ha of available farm land remains unused, however. Working capital is constraining and so is labour during the months of April, May and December. Assuming higher level of variable inputs and the possibility of hiring labour during the above critical months, a maximum gross margin of 2,615 shs is obtained from a cropping pattern which utilises all available farmland, with an increased cultivation of 0.55 ha. maize-groundnut, 0.65 ha maize and 4.9 head of cattle. Working capital needs to be increased to 427 shs and hired labour in the amounts of 33 hours in February, 100 hours in April, 95 hours in May and 100 hours in December has to be secured. Altogether some 400 shs additional working capital is required, which makes this alternative - other than for the sake of increased physical production - not much better than the previous one. - 122 - 2.1.8 1-i!- Y-1 - ' ' t riing in Iringa The s':.rver arecl' Y r-ng covers the Divisional census areas 0423, 0425, 0428, 02. 0 ~ E.0'^, P2 f--'rr are surveyed but complete records are available for -7 . . :-? rictics of these farms are summarised in Table 2.49. The Irince farmer in 46 years old and has attended school for 1 year only. Altogether 5.5 people live in the household representing 3.1 man-equivLlonts available for farm work. Differences between farms are small when farms are grouped by their levol of production and might relate to a slightly higher labour availability for the better farmers. Average farm size is 2.32 hectares. Although this does not include fallow and pasture land, only 83 percent of it is cultivated. A crop mixture bf maize, beans and some vegetables occupies more than half of the cultivated land. Maize and millet occupy a. further 36 percent, the rest being taken up by a mix of maize/groundnut and/or sunflower and beans or cassava and other crops. Best farmers are larger than average by about 43 percent. They cu1livate less of the crop mix maize/beans/vegetables and more of all other crops or crop mixes. Cattle keeping is of significant importance in Iringa; an average of 7.1 head of cattle are kept per farmer. But cattle seem to be concentrated in the hands of the better farmers and one could. say that cattle keeping makes these farmers better. Variable costs amount to 193 shs/farm, 89 percent of which is spent on seed. Records on labour inputs are available for crop production but missing for livestock. They average 780 hrs (336 hrs/ha). Higher labour inputs can be attributed to increases in cultivated area. Value of productinn averages 2,J52 shs per farm. This is made up of 37 percent returns from livestock and the remaing 63 percent from crop cultivation, in which productivity averages 847 shs/ha. Livestock enterprises make up more than half of the total value of production for better farms. But only crop cultivation indicates a steady rise in productivity from 292 to 906 shs/ha for farm groups from louett to highest. Labour productivity follows the same trend but the rise is more marked: from 93 to 805 shs per man-equivalent for the same farm groups. -122/a- SUMARY TABLE 2.49: Composite Characteristics of Iringa Farms - grouped by level of farm production - Details Very Leyel of farm production Averag Very IVery high. High Medium Low low No. of observations 15 15 15 15 15 1 75 Farmer, family, labour I Farmer's age, yearn 48.3 47.6 46.1 45.2 42.1 45-9 Farmer's education, school-yrs 0.8 1.1 1.0 0.9 1.6 1.0 Farm family, people 6.8 6.0 5.7 4.8 4.5 5.5 Perm. hired labour, nos. - - - - Total people on farm 6.8 6.0 5.7 4.8 4.5 5.5 Labour availability, man-equiv. 3.4 3.3 3.0 2.9 2.3 3.1 FARI LAND Farm wize, ha 3.31 2.90 2.54 1.84 1.00 2.32 (standard error (0.671 (0.27) (0.39) (0.32)1 (0.20) (0.20) Cropping pattern, % Maize/beans/vegetables 39.1 55.2 47.4 48.4 91.8 51.6 Maize millet 44.5 35.6 33.8 46.0 6.1 35.9 Maize/groundnut/sunflower 10.0 5.0 15.5 - 1.0 7.8 Beans, cassava, others 6.4 4.2 3.3 -5.6 1.1 i 4.7 TOTAL *100.0 100.0 100.0 100.0 100.0 100.0 Livestock, head of cattle 18.3 8.9 3.6 1.1 3.5 7.1 Variable inputs, shs 305 303 176 105 78 193 Percentage distribution: Seed 79,0 91.1 100.0 87.6 100.0 89.1 Fertilizer - - - Pesticides - - - - 8.3 Hired labour 14.4 7.9 - 12.4 -8.3 Hired -ower 6.6 10 -2.6 Others* - - - TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 Labour inputs, hrs 1,372 1,222 499 405 406 735 Perc-ntage distribution: Crop production 100.0 100.0 100.0 100.0 100.0 100.0 Livestock prouction n.a n.a na n.a n.a n,a TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 Value of farm production, shs 6,1221 3,351 1,846 1,069 377 2,552 (standard error) (699.1) 1(167.0) (99.9) (50.5) (36.5) (276.3) Percentage distribution: 7 Maize/beans/ve-etables 20.0 54.5 58.5 65.8 73.6 Maize, millet 21.1 17.5 16.1 19.2 3.9 20.0 Maize/groundnuts/sunflower 6.3 4.1 12.3 - - 6.3 Beans, cassava, others 1.6 0.2 0.9 - - 1.0 Livestock 51.0 23.7 12.2 15.0 22.5 37.0 TOTAL 100.0 100.0 1 100.0 100.0 100.0 100.0 - 122/b - SUMIARY TABLE 2.49 (cont'd) Level!of farm production Vive Detail Avrage Very Ve ry high High i Medium Low low Value offhrm production, shs 6,122 3,351 1,846 1f069 377 2,552 Variable costs, shs 305 303- 176 1l5 78 193 Sum of gross margin, shs 5,817 3,048 1,670 1 964 299 2,359 (farm income) Off-farm income, shs 1,986 432 60 333 1 347 632 Family income, shs 7,803 3,480 1,730 1,297 646 2,991 Value of home consumption, sh n.a n.a n.a n.a n.a n.a Household expenditures, shs 1,028 1,373 693 1,635 1,299 1,173 Cash for investment, replacement, savings, consumption 6,775 2,107 1t037 -338 -653 1,818 LAND PRODUCTIVITY (crops) Value of production, shs/ha 906 882 638 493 292 1 847 Variable costs, shs/ha 92 104 69 8 83 Gross marEin, shs/ha 874 778 569 436 214 764 Labour input, hrs/ha 414 421 196 220 406 336 LABOUR PRODUCTIVITY (crops) Value added, shs/man-dquiv. 805 691 481 281 93 I 483 Value added, shs/labour hr. used 1.99 1.86 2.89 2.01 0.53 1.92 Ratios Value of output to value of input, a. Labour not costed 20.1 11.0 105 10.2 4.8 13.2 b. Labour costed 4.6 1.4 3.7 2.7 0.9 2.7 Degree of family labour use, % (crops only) 27 25 11 9 12 17 Degree of commercialisation, % noa n.a n.a n.a n.a n.a Farm income as percdnt of total income 73 87 96 74 46 79 Cash balance as percent of total income 1/ 86 60 59 negative 61 1/ Including value of home consumption - 123 - Farm income averages 2,359 shs and is spread from 299 to 5,817 over the farm groups. Off-farm income(accruing mainly to the best farm group) averages 632 shs per farm. Family income thus averages 2,991 shs per farm family, spreading from 646 to 7,803 shs over the farm groups. Household expenditures average 1,173 shs spread evenly over all farms. The value of home consumption could not be calculated for the Iringa area. Thus, the balance to mect home consumption and possible investment, replacement, or savings averages 1,818 shs per farm. Comparable values from other areas indicate that this amount would be sufficient to meet home consumption requirements. However, the balances are negative for the two lowest farm groups and only sizeably bigger than average for the best groups. 21 percent of all Iringa farmers are engaged in off-farm employment. ThErwork an average of 173 days per year at 18.74 shs/day. This, spread over all Iringa farmers, yields an off-farm income of 632 shs per farm. The farm size in Iringa averages 2.32 hectares, of which 1.92 hectares are recorded to be-under crops. The farm size structure indicates unqqual distribution of land whereby a large proportion of farmers have below average farm sizes while a few privileged ones hold disproportionately much of the total land. Farm size group, ha Average farm Percent of Percent of total size, ha farmers area Under 1.00 065 25.3 7.1 1.00 - 1.99 1.52 26.7 17.5 2.00 - 2.99 2.53 24.0 26.2 3.00 - 3.99 3.37 8.o 11.6 4.00 - 4.99 4.43 9.3 17.8 5.00 and above 6-85 6.7 19.7 2.45 100.0 100.0 The cropping pattern is shown below: Crop, crop mix Average cultivation Percent of total size, ha Maize/beans/vegetables 0.99 51.6 Maize, millet 0.69 35.9 Maize/groundnuts/sunflower 0.15 7.8 Beans, cassava, others f 0.09 4.7 Maize/br',nsvetables A mixture of maize and beans and/or vegetables is grown by all but six farmers in the area. Farmers cultivated an average of two fields, each 0.66 ha. The joint value of production of this crop mix averages 867 shs/ha. and after deduction of 124 shs in variable costs (95 percent for sedd), the gross margin amounts to 723 shs/ha. The grouping of the production coefficients according to level of productivity suggests that there exists an inverse relationship between cultivation size on the one hand and seed rate, labour inputs and productivity on the other. The value of production is alimost triple on half the average cultivated area, with about double the average seed rate and labour inputs. Cobb-Douglas production funttion analysis indicates that production elasticities are high for land (0.44*) and seed (0.32*). Laboun in land preparation only yields an elasticity of 0.10 and labour in weed control a negative elasticity (-0.02). The sum of elasticities amounts to 0.84 and the adjusted R2 is 0.48. Marginal productivities are higher than factor costs for both land and seed rate. This suggests that, at the opportunity costs assumed, acreage should be expanded and seed rates increased, while nroportionately less labour should be applied. Limiting the optimum combination to the present cost level of 297 shs (including opportunity costs for labour and land) an input of 0.88 hectares, 104 shs worth of seed and 35 labour hours in land preparation could yield 599 shs as return. - 12 - TPBLE_2.30: Input/Output Coefficients by Level of Productivity - Maize, beans, vegetables, Iringa - -nput/ouitput coefficients LAverage Very Very high High Medium2 Low low No. of observations 21 21 20 20 20 102 O-i.tivation size, ha 0.33 0.42 0.68 1._02 10.72 0.66 Variable inouts, shs/ha 5 ud 201 181 147 85 94 118 hired labour - - - 10 11 5 Hired power 5 - - . 1 .T -? 7A1:201 186 147 95 105 124 ts hrs/ha Land preparation, planting 268 241 116 103 146 153 Weed control 96 105 1 65 45 70 68 BHarvesting, others 240 211 122 6 10 122 TOTAL 604 557 303 244 323 343 Value of production shs/ha 2,553 1,393 934 705 428 867 Variable costs, shs ha 201 186 147 95 105 124 Gross margin, shs/ha 2,352 1,207 787 610 323 723 Maize Maize in purestand is grown by 39 of the 82 surveyed Iringa farmers. Because some farmers have more than one maize field, 48 observations are available on maize cultivation. Cultivation sizes average 1.24 hectare. Yield amounts to 2,917 kg/ha, which at a producer price of 0.75 shs/kg, yield a value of production of 2,187 shs/ha. Variable inputs relate to seed (17 shs), hired labour (19 shs) and power (7 shs) and total 43 shs/ha. They are equally low in all productivity groups. Also labour input levels can be observed in comparing the productivity groups (Table 2.51). TAIBLE 2,51: ofv/On~0~> Prori-l Ay -02ze, ~na Input/output coefficients of Average cry Very ___ .gh Ne--um Low I low No. of observationst6 6 7 7 7 48 Ciltivation size, ha nr. o39 0 97 n.o3 1.28 1.24 Variable nuts, shs/ha Seed 18 17 15 15 16 17 Hired labour - 27 -21 36 i 19 Hired power 1- - 7 TOTAL 18 63 15 36 52 43 abour inputs, hrs/ha Land preparation, planting 67 23 93 74 5o 63 Weed control 52 22 38 48 40 43 Harvesting, others '8 1 22 A6 43 4 TOTAL 117 67 77 166 130 150 Yield, kg/ha 7,200 5,891 2,533 2,009 1,180 2,917 Value of production, shs/ha 5,400 4,418 1,899 1,507 856 2,188 Variable costs, shs/ha 18 63 15 36 52 43 Gross margin, shs/ha 5,382 4,355 1,884 1,471 804 2,145 Production function analysis produced positive elasticities for land, seed and labour in weed control, of 0-53*, e.20 and 0.14 respectively. Labour in land preparation has a negative elasticity of -0.21. Their sum is only 0.66 at an adjusted 2 R of 0*58- Marginal productivities are similarly high for land, seed and labour in weed control. For all three inputs they are higher than opportunity costs, suggesting that for the optimum solution they should be proportionally increased. Restricting their costs to present total costs, maize should. be grown on fields of just one hectare, the seed rate should be quadrupled while labour inputs should remain at present levels. This combination would result in a 46 percent increase in productivity (Shs 3,189/ha). Millet Millet is grown by 9 farmers. Complete records are available from 6 of them, who grew it on 0.49 ha on average. The value of seed is 52 shs, and labour inputs average 237 hrs. Yields are 482 kg/ha. Producer prices were very high and averaged 4.43 shs/kg. Thus, the value of production averaged 2,136 shs and gross margins 2,084 shs/ha. Higher productivity is achieved on smaller than average acreages with both higher seed rate and labour inputs (Table 2-52). - 126 - TABLE 2.52 Input/Output Coefficients by Level of Productivity - Millet, Iringa - Level of productivity Input/output coefficients Average High Low No. of observations 3 3 6 Cultivation size, ha 0.32 0.91 0.49 Variable inputs, shs/ha Seed 87 37 52 Labour inputs, hrs/ha Land preparation, planting 201 77 122 Weed control 78 35 44 Harvesting, other 115 60 72 TOTAL 494 112 23 Yield, kg/ha - - i 482 I Value of production, shs/ha 6,354 819 2,136 Variable costs, shs/ha 87 37 52 Gross margin, shs/ha i 6,267 I 782 I 2,084 Maize/groundaut s This crop mix is grown by 12 farmers, on 0.54 hectares each. Nine complete records are available. Inputs for seed average 164 shs/ha and labour inputs 394 hrs/ha. For this average input level, value of production amounts to 1,620 shs and gross margin to 1,456 shs/ha. (Table 2.53). As for the other crops or crop mixes, move intensive cultivation (higher seed rates, higher labour inputs) results in significantly higher levels of productivity. Value of production as well as gross margins can be tripled. Minor crops or crop mixes In Iringa a variety %,f crop or crop mixes are grown, but to a very limited extent. However, some of them indicate high levels of productivity. Table 2.54 gives summary input/output information relating toa number of minor crops. - 127 - TABLE 2.53 Input/Output coefficients by Level of Productivity Input/output coefficients Level- rodutivity Average __High Low No. of observations 5 9 Cultivation size, ha 0.21 0.54 0.54 Variable inputs, shs/ha Seed 342 119 164 Labour inputs, hrs/ha Land preparation, planting 234 91 143 Weed control 220 57 87 Harvesting, others 294 .110 164 TOTAL 749 258 394 Value of production,shs/ha 4,482 835 1,620 Variable costs, shs/ha 342 119 164 Gross margin, shs/ha 4,140 716 1,456 Livestock Livestock keeping is of great importance in Iringa. 37 percent of the total value-of farm production is derived from livestock keeping. 21 farmers kept cattle, at an average of 14 head each. Three of the cattle keepers had sheep and 5 had goats. 24 farmers kept poultry, at an average of 4 birds per farm (range 1-16). Whole farm The composite value of farm production from both crop cultivation and livestock keeping averages 2,552 shs for the Iringa farms. 37 percent comes from livestock keeping, which also accounts for 51 percent for the best farm group. As mentioned before, livestock keeping might be saidto make the best farms, as the difference between this group and the next best seems to be due only to the number of cattle kept. Labour inputs relate to crop production only and average 780 hrs for the average farm of 2.32 ha. Variable costs of an average of 193 shs/farm deducted leaves the sum of gross margin of 2,359 shs, ranging from 299 to 5,817 over the farm groups (Table 2.55). - 128 - TABLE 2.54 Input/Output Coefficients by Level of Productivity - Minor crops, Iringa - Input/Output Coefficients Bananas Beans/ Maize/ Sunflower Vegetables vegetables sunflower No. of observations 3r 4 3 _ 3 . Cultivation size, ha0.14 0.39 0.52 0.23 0.13 Variable inputs, shs/ha Seed 4 38 28 160 33 Hired labour - 14 10 - - TOTAL 4 38 160 33 Labour inputs, hrs/ha Land preparation, planting 107 71 95 380 420 Weed control 97 29 80 250 583 Harvesting, others 49 47 180 500 325 TOTAL 253 148 355 1130 1,328 Ja- - 16 - - Feb- 11. 80 380 - Mar- - - - 175 Apr 117 - - 250 495 May - - - - 225 Jun - 63 102 -45 Jul- 11 77 - 63 Aug 28 - 500 38 Sep- 18 - - 288 Oct---- Nov - - - Dec 845 80 TOTAL 148 355 1,130 1,328 Family 253 137 345 1,130 1,328 Seas. labour - 11 10 - - TOTAL 253 148 355 1,130 1,328 Yield, kg/ha - - 1,167 - Value of productiony shs/ha 1,732 905 2,036 2,833 3,500 Variable costs, shs/ha 4 52 38 160 33 Gross margin, shs/ha 172878535 198 2,673 z3,467 0/ - 129 - This table also groups farms according to farm size and it can be seen that as farm size increases, the number of cattle kept, labour input (crops only), and variable inputs increase. Proportionate increases are also realised in the value of farm production and gross margin (farm income). When grouping is done according to farm size, productivity decreases sharply with increase in cultivation size 'and amounts in the largest farms to only 40 percent of the productivity achieved in the smallest farms. As pointed out for other survey areas, it is the slightly larger than average farm which yields maximum returns through increased area coupled with high productivity. Cobb-Douglas function produced positive elasticities for all parameters included, and especially high elasticities for variable inputs (0.40*) and land (0.24*). For labour and livestock numbers elasticities of 0.12 and 0.16 are obtained, bringing the sum of elasticities to 0.92 with adjusted R2 of 0.47. Marginal productivities are higher than likely factor costs for variable inputs and land, suggesting that use of these factors, if possible, should be increased. An attempt to optimize resource combination yields 1,670 shs/ha, which is 50 percent higher than at present. This holds true for only the asuumed factor costs. Farmers.' attitudes Above everything else, Iringa farmers consider the price offered for both maize and beans insufficient. In addition, they indicate that transportation, especially for maize, poses serious problems to them. Other opinions, especially with regard to prices and availability of material inputs, such as seed and fertilizer, are only sporadically expressed. Potontial enterprises for Iringa region could be groundnuts, millet, cowpeas, sunflower and wheat. The reasons expressed as to why these enterprises are not presently extensively cultivated are diverse and relate to a combination of insufficient supply of resources (land, labour, seed) and low producer prices. If farmers would have a completely free hand in deciding on enterprise combinations, without constraints, they would grow the following crops in the following proportions: Maize 3.4 ha Beans 1.1 ha Groundnuts 0.2 ha Cowpea 0.1 ha Sunflower 0.1 ha Others 0.2 ha 5.1 ha - 130 - TABLE 2.55 (a) Input/Output Coefficients for Farms by Levels of Farm Production - Whole farm, Iringa - Level of production Input/output coefficients Leve of Average Very V ery high High Medium Low low No. of observations 15 15 15 15 i 15 75 : Cultivation size, ha 3.1 i 2.90 2.54 1.84 1.00 2.32 Cattlo, head 18.3 8.9 3.0 1.1 3.5 7.1 Labour inputs, (crops only) hrs : 1,372 1,222 499 45 406 780 Value of production, shs 6,122 3,351 1,846 1,069 377 2,552 Variable inputs, shs 305 303 176 105 78 193 Gross margin, shs 5,817 3,048 1,670 299 2,359 (b) Input/Output Coefficients of Farms by Farm Size Classes - Whole farm, Iringa - Farm size group, ha Input/Output coefficients Under . 1.00 - 2.00 - 3.00 - 4.00- 500 & 1.00 1.99 2.99 3.99 4 99 above iNo. of observations 19 20 18 6 7_ 5 Cultivation size, ha 0.65 1.52 2.53 3.37 4.43 6.85 1.9 4.2 8.8 2.7 16.0 24.6 Labour inputs, hrs 500 443 1,112 534 1,112 1,838 Value of production, shs 1,426 1,880 2,771 3,280 3,747 67194 iVariable costs, shs 65 126 261 196 342 499 Gross margin, shs 1,361 1, 75- 2,510 3,084 3,405 5,695 _ _ _ __A - 131 - TABLE 2.r6 (a) Farm and Failly Income and Household Expenditures -Iringas sh) . Level of farm production IDetails of income and Average expenditures Very Very hiRh High Medium Low low Value of farm production, sha 16,122 3,351 1,846 1,069 377 2,552 Variable costs, shs 305 303 176 105 78 193 Sum of gross margin, shs 5,817 3,048 1,670 964 299 2,359 (farm income) 0ff-farm income, shs 1,986 432 60 333 347 632 Family income, shs 7,803 3,480 1,730 1,297 646 2,991 Value of home consumption, shs not applicable ,Household expenditures, sha 1,028 1,373 693 1,635 1,299 1,173 Cash for investment, replacement, savings, consumption 6,775 2,107 1,037 -338 -653 1,818 (b) Composition of Household Expenditures Item Shs Percentage of total Clothing 616 52.5 Household goods 202 17,2 Detergents 130 11.1 Condiments 70 6.0 Sugar 56 4.8 Kerosine 43 3.7 Alcoholic drinks 16 1,4 Cooking oil 11 0.9 Tea 8 0.7 Others 21 1.7 TOTAL 1,173 100.0 - 132 - Crops are mentioned in purestand. The acreage would,be increased by 50 percent over the present farm size. Nearly all farmers would keep cattle, on average 31 head per farmer, which seems eimely ambitious. Finally, farmers expressed their views as to how to use additional income should it be forthcoming: Preference, percent 1st 2nd 3rd Build house 76 15 7 Investment in agric. 10 3 3 Consumption 14 76 56 Save 3 20 Others 3 14 First and foremost, new houses are desired. This requirement satisfied, farmers would prefer to use additional income to increase consumption. Also savings is considered. The desire to invest these savings either in agriculture or in other industries seem to be very low. Farm proLramming Iringa farms are programmed using the following activities (enterprises) in both their average and best productivities: maize-beans-vegetables, maize, millet and maize- groundnut. In addition, livestock keeping is included. Constraints relate to the uverage farm size of 2.32 ha, average level of working capital of 193 shs/farm and a family labour supply of 387 hours/month. The resource combination yielding maximum farm incomes includes 1.27 ha maize and 1.05 millet, both in their best productivity. In addition, 4.2 head of cattle are kept. Farm income would amount to as much as 14,816 shs. Labour in February, April and June constrains further output. Farm programming for Iringa is to some extent unsatisfactory. Potential increases in returns are indicated to be extremely high, but this is due to the high performance of both maize and millet in their pest productivy. Available data, however, cannot fully explain the reasons which underline such high productivities. 133 2.1.9 Cereal and Tobacco Farming in Ruiuma The Ruvuma survey area is delineated by the divisional.census areas 12111 1222 and 1225. Results are available from only 29 farmers, whose farming characteristies are summarized in Table 2.57. As for other areas, the 29 farms are stratified into 5 equal groups according to their level of farm production, although the results of this endeavour turned out not to be conclusive. The Ruvuma farmer is on average 42 years of age and has, relative to other areas, a high level of educational standarO (3.9 school-years). The farm family comprises 5.4 people, who form a potential work force of 3.0 man-equivalents. The cultivated average farm size of 3.24 ha. on average is quite large. The crops grown are mainly cereals (maize, Millet), pulses, groundnuts and tobacco as cash crop. Best farmers cultivate 50 percent more acreage than average, with bigger acreages devoted to cereals and less to pulses and oilcrops. Livestock are kept by few people only; cattle average 1.3 head over all farms. The average value of production amounts to 4,575 she per farm and is doubled for best farmers. Variable costs amount to 572 shs per farm on average, but are slightly less than this for better farmers. Labour inputs amount to 3,217 hours per farm, which represents 73 percent of the family's work potential. Farm income averages 4,003 shs per farm and is more than double this amount for best farmers. Off-farm incomes seem to be equally spread over all farm groups and together with farm income yield, a family income of 4,471 shs. Since the amount of home consumption was not recorded the cash balance of (after deducting 2,183 shs in household expenditures) 2,289 shs includes the value of subsistence production. Generally, the data available for Ruvuma are relatively thin. For this reason their analysis does not yield the same level of reliability as achieved elsewhere. - 134 - TABLE 2.57 Composite Characteristics of Ruvuma Farms - grouped by level of production - Level of productioh Details Vr eAverage Very Vr high High Medium Low. low Number of observations 5 6 6 6 6 29 Farner, family, labour 9 Fatmersi age, years 40.8 43.3 45.2 47.8 33.8 42.2 Farmers' educationI school-yre 3.8 4.0 2.2 4.0 5.8 3.9 Farm family, people .5.6 5.1 6.2 6.0 4.5 5.4 Per. hired labourers - 0.1 - - - Total people on the farm 5.6 5.2 6.2 6.0 4.5 5.4 Labour availability, man-equiv. 3.151 2.58 3.50 [ 3.46 2.5 3.0 FAW LAND Farm size, ha 4.56 i 3.40 2.45 4.02 1.99 3.24 (standard error) (1.91 (0.46) (0.22) (1.80) (0.37) (0.5m) Cropping pattern, %o Cereals 68.1 46.0 31.7 63.5 49.0 53.7 Pulsed, oil crops 5.1 16:6 - 20:5 28.6 13.8 Maize, beans 13.2 29.8 57.7 13.1 18.6 24.2 Tobacco 4.6 7.5 10.6 2.9 - 5.2 others 8.9 - - - 3.8 3.1 TOTAL- 100.0 100.0 100.0 100.0 100.0 100.0 Livestock, head of cattle - 5.2 0.6 -0.5 1.3 Variable inputs, she 455 493 539 928 426 572 Percentage distribution: Seed 70.4 32.0 17.7 17.2 26.. 27.9 Fertilizer 14.8 10.0 22.3 8.9 13.4 12.5 Pesticides 3.0 - 0.9 - 1.5 0.7 Hired labour - - 8.2 56.8 19.4 1 272 Livestock exoenses 11.84 31.8 16.1 23.9 22.3 Others 11.8 14.6 19.1 1.0 1.7 9.4 TOTAL 100.0 1100.0 100.0 100.0 100.0 100.0 Labour input, hrs (not costed 3,889 3,053 3,255 2,799 3,203 3,217 Percentage distribution: crops 84 63 53 57 39 59 Livestock 16 37 47 43 61 TOTAL 100 100 1001 100 100 100 Value of farm production, shs 9,027 5,003 4,166 3,084 2,338 4,575 (standard error) (1892.76) (96.94) (162.60) (137.40) 1(132.31) (518.58) Percentage distribution: cereas 66.7 43.4 25.6 40.0 41.0 45.9 Pulses, oil crops 1.9 10.3 - 22.4 13.1 7.4 Maize, beans 5.8 25.0 1 54.8 22.4 28.5 23.9 Tobacco 2.4 10.2 15.2 6.8 - 6.8 Other crops 23.2 - - 8.4 1 14.2 11.0 Livestock - 11.1 4.4 -. 5.0 TOTAL 100.0 1100.0 100.0 100.0 100.0 100.0 i - 135 - SUMARY TABLE 2.57 (conttd) fDetails Very Level of farm Droduction Very Average high High Medium Low low Value of farm production, shs 9,027 f 5,003 4,166 3,084 2,338 4,575 Variable costs, shs 455 493 539 928 426 572 Sum of gross margin, shs 8572 4,510 3,627 2,056 1,902 4,003 (farm income) Off-farm income, shs 70 600 270 133 70 273 676 468 Family income, shs 9,272 5,110 3,897 2,189 2,578 4,471 Value of home consumption, she n.a n,a n.a n.a n.a n.a Household expenditures, shs 2,482 2,686 1,617 2,321 1,856 2,182 Cash for investment, replacement, savingg, consumption 6,790 2,424 2,280 722 2289 LAND PRODUCTIVITY Value of production, shs/ha 1,979 1,308 1,625 767 1,137 1,341 Variable costs, shs/ha 100 82 150 193 161 137 Gross margin, shs/ha 1,879 1,226 1,475 574 976 1,204 Labour input, hrs/ha 716 565 To4 392 586 LABOUR PRODUCTIVITY Value added, shs/man-equiv. 2,721 1t748 1,048 710 793 1,384 Value added, shs/labour hr used 2.20 147 1.13 62 1.29 Ratios Value of output to value of input a. labour not costed 2198 10.1 717 323 5.4 8.0 a. labour costed 2.1 14 1.1 0.9 0.7 1.2 Degree of family labour use, % 82 79 62 57 8673 Farm innome as percent of total income 92 88 93 94 74 89 - 136 - The cultivated farm sizes average 3,24 hectares. Values attached to the land are 600 shs/ha. The distribution of land seems to be more equal than in many of the other survey areas, as indicated below. Farm size group, ha Average farm size, Percent of farmers Percent of total ha area Under 1.00 0.89 3,3 0.9 1.00 - 1.99 1.43 24.1 10.6 2.00 - 2.99 2.50 37.9 29.3 3.00 - 3.99 3.55 20.7 22.7 4.00 - 4.99 4.77 6.9 10.2 5.00 and above 12.36 6.2 26.3 3.24 loo.0 100.0 In spite of the presence of both extremes of very small and very large farm sizes, the majority of farmers hold farm sizes not too far from the average of 3.24 ha. The cropping pattern includes a variety of crops, as shown below: Crop, crop mix Average cultivation Percent of total Ie size, ha Maize, beans, cassava 0.78 24.1 Maize 0.75 23.1 Maize, millet 0.66 20.4 Sorghum 0.20 6.2 Other cereals, (millet, rice) 0.13 4.0 Cassava 0.36 11.1 Beans, groundnuts 0.09 2.8 Tobacco 0.17 5.2 Others,(cotton, coffee, vegetables) 0.10 3.1 TOTAL 3.24 100.0 Maize, beans, cassava A crop mix of maize, beans and cassava is most important in the cropping pattern and is cultivz:ted by 27 farmers on 1 hectare each. Average field sizes, since some farmers cultivate this mix on more than one field, average 0.70 ha. The value of production of the crop mix averages 1,740 shs/ha. In Table 2.58 the results of this crop mix are grouped according to productivity and fertilizer application. 0/* - 137 - Higher productivity coincides with smaller cultivation sizes, higher material input and higher labour inputs. The higher yields of fertilized fields is due to higher variable and labour inputs. The returns to additional labour used on fertilized fields amounts to 1.52 shs/hr. Cobb-Douglas function analysis produced for all variables high marginal products, except for labour in land preparation. This suggest that all variables (except the last mentioned) should be increased proportionately. The respective elasticities are: land 040, seed 0.46*, fertilizer 0.03, labour in land preparation -0.36, and labour in weed control 0.52*. The sum of elasticities is 1.09 with an adjusted R2 of 0.35 Marginal productivities for all factors with positive elasticities are higher than assumed factor costs. Optimal re-allocation of factors is indicated as follows: land 0.95 ha, seed 69 shs, fertilizer 4.50 shs and labour for weed control 75 hrs. At the same tbsolute input levels and under the assumption that no constraints for the expansion of individual resources exist, farmers could adjust their input combinations not only to improve returns absolutely (3,060 shs), but also to increase their productivity bSy 85 percent. But again, these results should be treated as only indicative. Maize Maize in purestand is grown by 12 farmers or 40 percent of all farmers on cultivation sizesof 1.19 ha on average. Yields average 2,357 kg/ha and values of production at an average producer price of 0.59 shs/kg, 1,387 shs/ha. Only two farmers had fertilized their fiel s, but with no better results than the rest. Purchased inputs are low and amount on average to only 83 shs, of which 15 shs is for sedd, 9 shs for fertilizer and the rest mainly for hired labour. Labour inputs average 423 hrs/ha, 174 hrs for land preparation and planting, 101 hours for weed control ane the rest for harvesting. Due to the limited number of observations, the data on purestand maize are stratified only according to high and low performance. TABLE 2.58 Input/output Coefficients Grouped by Level of Productivity and Fertilizer Use, - Maize, Beans, Cassava v Ruvuma - Average.. ised ised Input/Output Coefficients Level of productivity Fertil- Unfertil- high High Medium Low low L No. of observations 6 5 5 5 5 26 3 23 Cultivation size, ha j 0-46 0.45 0.76 0.82 1.03 0.70 0.73 0.69 Var-able inputs, shs/ha Seed 46 31 12 19 25 27 16 27 Fertilizer 14 - - 13 Others 56 73 13 27 -39 49 25 TOTAL 102 128 58 46 25 66 138 52 Labour inputs, hrs/ha Land preparation, planting 321 272 184; 291 290 272 254 274 Weeding 234 207 161 158 132 196 395 169 Harvesting, others 2 375 20 2o3 142 190 160 1 TOTAL 82 854 545 652 564 658 809 637 Value of production, shs/ha 2,830 2,224 2,037 1,663 746 17 2,045 1,698 Variable costs, shs/ha 102 128 58 46 25 66 138 52 Gross margin, shs/ha 2,728 2,096 1,979 1,617721 1,674 1,907 1,646 721 1674_ 1,907 46e - 139 - TABLE 2.59 Input/Output Coefficients b2 Level of Productivity - Maize, Ruvuma - Input/Output coefficients High Low Average No. of observations 6 6 12 Cultivation size, ha 0.62 1.75 1.19 Labdur input, hrs/ha Land preparation, planting 206 183 174 Weed control 150 11 101 Harvesting, others 165 192 148 TOTAL 521 446 423 Yield, kg/ha 4,536 2,02 2,357 Value of production, shs/ha 2,776 1,181 1,387 Variable costs, shs/ha 23 79 83 Gross margin, shs/ha 2,753 1,102 1,304 Better results from maize cultivation are achieved on smaller acreages with higher labour inputs, especially in weeding. Variable inputs are even less than on average. This-is also confirmed by Cobb-Douglas Production Function Analysis. Only land and weeding labour have relatively high elasticities of 0.15 and 0.66 respectively. Variable costs have -0.14 and labour in land preparation 0.03, bringing the sum of elasticities to 0.70. Adjusted R2 is .63. Compared to their opportunity costs, ratios of 3.47 and 9.07 are calculated for both land and labour in weed control respectively. Optimal combination of resources would yield 87 percent higher produttivity for this crop by using, on the same acreage as at present, significantly more labour for weeding (313 hours) instead of land preparation. Maize,Millet A crop mix of maize with millet is grown by 9 farmers on 12 fields of 1.33 ha each. The joint yield of both orpps average 2,803 kg/ha. At producer prices of 0.54 shs/kg the value of production a mounts to 1,552 shs and is much higher than both crops grown in purestand. Material inputs are low and amount to 36 shs/ha only, of which 32 -hs are for seed. Labour inputs average 757 hrs/ha, which is more than for maize in purestand. - 140 - Since observations are few, data for this crop mix have also been stratified only according to high and low performance. (Table 2.60). TABLE 2.60: Input/Output Coefficients by Level of Productivity -- Maize, Millet, Ruvuma - Input/Output coefficients High Low Average No. of observations 4 59 Cultivation size, ha 0.61 2.04 1.33 Labour inDut, hrs/ha Land preparation, planting 267 232 241 Weeding 514 295 385 Harvesting, others 323 178 171 TOTAL 1,104 705 79? Yield, kg/ha 3935 2,483 2,803 Value of production, shs/ha 29125 1,341 1,552 Variable costs, shs/ha 31 32 32 Grofs margin, shs/ha 2,094 1,309 1,520 Yields and values of production are considerably higher on the better performing farms. As already found for other crops in this area, this can be accounted for by higher labour inputs for weed control on smaller cultivation sizes. This is basically confirmed by Cobb-Douglas production function analysis. Land has a high elasticity of 1,20, labour in wee& control 0.48, seed 0.17 and labour in land preparation -0.97. Their sum is 0.38, and the adjusted R2 is 0.92. A high marginal product was especially calculated for land. Due to the extremely high marginal product-to-opportunity-cost ratio for land, the effects of the other inputs are overshadowed and an attempt to optimize resource combination would yield results which are out of proportion. It can safely be said, however, that increased rates of seed, increased labour inputs for weed control and larger cultivation sizes can contribute to raising production and returns of this crop mix. - 141 - Tobacco Seven farmers grew tobacco on 0.58 ha each. In the rotation tobacco seems to rotate with only maize or with itself. Yields amount to 542 kg/ha which, at average producer prices of 4.16 shs/kg, produce a value of production of 2,255 shs/ha. Ninety percent of all tobacco is fertilized. The unfertilized observation shows the lowest yield and value of production (half of the average). Total variable costs amount to 146 shs/ha, 27 shs spent on seed, 97 shs on fertilizer and the rest for seasonally hired labour. Thus gross margins are calculated at 2,109 shs/ha. This is higher than for any other crop and it remains to be investigated why tobacco is not cultivated more extensively. Again observations are only stratified into a high and low farm group and the results are compared below. TABLE 2.61: Input/Output Coefficients by Level of Productivity - Tobacco, Ruvuma - Input/output coefficients High Low Average No. of bbservations 3 4 Cultivation size, ha 0.65 0.52 0.58 Variable inputs, shs/ha Seed 21 33 27 Fertilizer 107 87 97 Others - 22 22 TOTAL 128 142 146 Labour inputs, hrs/ha Land preparation, planting 215 426 324 Wedding 539 298 486 Harvesting, 6thers 860 t 804 812 TOTAL 1,614 1,628 1,622 Yield, kg/ha 693 399 542 iValue of production, shs/ha 2,883 1,663 2,255 Variable costs, shs/ha 128 142 146 Gross margin, shs/ha 2,755 1,521 2,109 - 142 - Cobb-Douglas production function analysis yielded high elasticities for land (0.61) and labour in weed control (0.56), while fertilizer levels used under the present mode of application seems to be too high (0.01). As for the previous crop mix, the high marginal productivity of land tends to overshadow possible effects of other inputs on yields. Ninor crops A number of 'minor' crops or crop mixes are also cultivated in Ruvuma. These include cassava, coffee, beans, rice, millet, sorghum, groundnuts, cotton, sugarcane, vegetables and orchards. Their input/output parameters are summarized in Table 2.62. Livestock 70 percent of all farmers in Ruvuma keep livestock of one sort or another. 17 percent keep cattle, on average 8 head (range: 3-15) valued at 452 shs per animal. Their productivity is not knoim. Average labour requirements range from 2,000 to 2,500 hrs/year. Half of the cattle keepers have alro goats, averaging 8 animals per farm. 50 percent of all farmers keep goats, on average 10 animals each (range: 2-23). Goats are valued at 69 shs each. Only two farmers kept sheep, 3 head on avergge and valued at 73 shs per animal. 17 percent of all Ruvuma farms keep poultry. The flocks are quite larger amounting to 23 birds on average (range: 6-67). Each bir0 is valued at 10 shs. Livestock seem to be an important component of Ruvuma farms, Unfortunately, input/output parameters are not available for their detailed analysis. Whole farm The value of farm production for the Ruvma farm amounts to 4,575 shs on average. This includes both produce sold and that retained for home consumption. However, livest-ck inventory changes, as well as livestock products, are not included. The value of production is mostly made up of cereals (46%) and a mixture of maize, beans, Cassava (24%). Other crops are of importance only in individual farm groups. In - 143 - Table 2.63 farms are grouped according to their level of farm production, as well as their sizes. From both methods of treatment, it can be seen that farm incomes, although having extremes at both ends, are rather clustered around the average. This also holds true for other parameters such as variable costs, labour input and cultivation sizes. Only farms with the highest level of production are distinctly different from the rest. Their values of production are double the average although having the aame level of variable inputs and only slightly higher labour inputs. However, farm sizes are considerably larger. Assuming constant returns to scale in these farms, 40 percent of increased farm production is due to increases in productivity rather than to scale. If this figure is applied to the average, it would raise farm production from 4,575 to 6,414 she per farm, which is realistically achievable by adopting the methods used by the better farmers in the same area. An attempt to use Cobb-Douglas production function analysis on the Ruvuma whole-farm data proved fruitless. Although land emerged with a high elasticity of 0.80, all other parameters included in the analyses had negative elasticities. In Table 2.64, the income position of the Ruvuma farmer is reviewed. Farm incomes average 4,003 shs, and are more than double this figure for the best performing farms. Even below average farms have higher incomes than are earned in many of the other survey areas* Off-farm incomes are rather evenly distributed among all farm groups and amount on average to 468 shs per farm. Family incomes thus average 4,471 shs, ranging from 2,578 to 9,272 she. Unfortunately, the portion of the farm income which goes into subsistence consumption cannot be specified. Consumer goods expenditures average 2,182 shs, leaving a balance which inciludes the value of home consumption as well as any cash accruing for investment or saving, of 2,289 shs per farm. Inputing a comparable level of home consumption as observed in other survey areas, Ruvuma farms may, on average, break even; it is doubtful, however, whether this holds true for the below- average farms. TABLE 2.62 Input/Output Coefficients for Various Minor Crops - Ruvuma - Details Cassava Coffee Beans Rice Millet Sorghum Groundnuts No. of observations 11 4 4 4 3 3 2 Cultivation size ha 0.68 0.40 0.65 033 0.54 1.63 0.17 Variable inputs, shs/ha Seed 79 - 41 84 13 9 153 Others 6 4040 27 41 207 - TOTAL 87 404 41 111 54 216 153 Labour input, hrs/ha Land preparation, planting 254 - 197 444 337 290 212 Weeding 123 255 46 1,023 299 132 280 - Harv.esting, other 173 1,431 60 1,060 252 159 279 TOTAL 550 1,686 303 2,522 888 581 771 Yield, kg/ha 2,275 1,825 377 1,203 1,497 718 500 Producer price, shslkg 0.66 .0,48 1 1.64 4.97 0,75 2.09 3.24 Value of production, shs/ha 1,503 6,352 616 5,977 1,122 1,498 1,618 Variable costs, shs/ha 87 404 41 111 54 216 * 153 Gross margin, shs/ha 1,416 5,748 575 5,866 1,068 1,282 1,465 * * Crops with only one observation have not been included. - 145 - RABLE 2.63 (a) Input/Output Coefficients for Farms by Level of Production - Whole farr, Ruvuma - I Level of production Input/output coefficients LvloprdcinAverage Very Very High High Medium Low low . No. of observations 5- 6 6 6 6 25 Cultivation size, ha 4.56 340 2.45 4.02 1-99 3.2! Head of cattle, no. i -5.2 0.6 - 0.5 1.3 Labour inputs, hrs/ha 3,25 Crop production, hrs 3,253 1,930 1,735 1,607 1,240 1,908 Livestock, hrs 636 1,123 1,520 1192 1 1,309 TOTAL 3,889 3,053 3,255 2,799 3,203 3,217 Value of production, shs/ha 9,027 5,003 4,166 3,o84 2,338 4,575 Variable costs, shs/ha 455 493 539 928 426 572 Gross margin, shs/ha 8,572 4,510 3,627 2,056 1,902 4,003 (b) Input/Output Coefficients for Farms Grouped by Farm Size - While farm, Ruvuma - input/output coefficientsFarm size group, ha_, coefficients Uder 1.CO - 2.00 - 3.co - 4. 00- 5.00 & 1.00 1.29 2.99 3.99 4.99 above No. of observations 1 7 11 _6 _ _2 _2 Cultivation size, ha 0 . .43 2 Cattle, head 3.0 - 1.5 2.5 1.5 - Labou;r-inputs, hrs Crops 1,545 1,123 1,749 2,177 2,106 4,711 Livestock, hrs 1,950 1,200 921 1,041 2,280 3,330 TOTAL 3,495 2,323 2,670 3,218 4,396 8,041 Value of production, shs 1,779 3,375 4,557 4",556 5,160 9,750 Variable costs, shs 342 450 493 410 601 002 Gross margin, shs 1,437 2,925 4,064 4,146 41559 7,748 - 146 - TABLE 2.64 (a) Farm abd Family Income and Household Expenditures - Ruvuma (shs) - Details of income and Level of farm production Average expenditures Vy e: e7 Very high High Medium Low low Gross margin (farra income) 8,572 4,510 3,627 2,056 1,902 4,003 Off-farm income 700 j 600 270 133 676 468 Family income 9,272 5,110 3,897 2,189 2,578 4,471 Household expenditures 2,482 2,686 1,617 2,321 1,856 2182 Value of consumption and cash for investment, replacement 6,790 2,z24 2t280 -132 722 2,289 and consumption (b) Cmmposition of Household Expenditures Items Shs Percentage of total Clothing 831 38.1 Household goods 229 10.5 Sugar 214 9.8 Detergents 181 8.3 Cooking oil 167 7.7 Medicine 159 7.3 Kerosine 121 5*5 Alcoholic Drinks 107 4.9 Condiments 96 4.4 Stimulants 17 0.7 Others 60 2.8 TOTAL 2,182 100.0 -147 - Farmerst Attitudes and Felt Constraints Like farmers elsewhere, Rumuva farmers have complaints and opinions about farming opportunities in their area. All farmers regard the producer price offered for maize as too low. In fact 0.59 shs/kg was considerably lower than for many other survey areas. Also the price of tobacco was regarded as too low by 23 per6ent of farmers, while 12 percent felt similarly for coffee prices. 57 percent of the farmers see difficulties in marketing beans. Insufficient oarket outlets are also mentioned for cassava by 30 percent and for millet by 17 percent of all farmers. With regard to the input situttion, all farmers query the absence of pesticides from the market. 57 percent also feel the same about hybrid or other improved seed. With regard to the fertilizer situation, the opinions of farmers are mixed: while 40 percent of them complain about its unavailability, another 40 percent question its cost. The same is true for machinery use, although such views are held only by 14 percent of all farmers. The reaction indicated by farmers to changes in producer prices is as follows: Acreages, ha Cro- No. of farms at high producer at low producer prices prices Maize 25 4.0 1.6 Beans 8 1.8 0.4 Cassava 4 2.0 0.7 Tobacco 7 2.7 0.6 Sorghum 3 2.5 0.1 Coffee 1.8 0.4 The ranges in cultivation sizes at high and low producer prices seem narrower for staple foods such as maize and cassava, indicating that a minimum of such crops will be growTn whatever the price situation. 0/. - 148 - Farmers in Ruvuma do not visualise alter!atives to the range of crops and types of livestock already held in the area. All constraints removed, Rumuva farmers indicated that they would grow the following enterprises in the following proportions (averages of all farms). Enterprise No. of Acreage, Percentage of farms a ha total Maize 27 4:09 48.3 Beans 18 1.57 18.6 Tobacco 7 0.80 905 Cassava 8 0.47 5.6 Millet 6 0..0 4.7 Coffee 4 0.38 4.5 Sorghum 2 0.34 4.0 Rice 5 0.33 3.9 Groundnuts 2 0.08 0.9 8.46 100.0 Apart from being very ambitious (more than twice present average acreages) this cropping pattern is very similar to the present farm organization. 83 percent of all farmers would, in addition, keep cattle and/or goats at a rather ambitious average number of 45 head. For the use of additional incomes the following preferences were expressed by the farmers. Preferences (percent) 1st 2nd 3rd Better housing 68 12 4 Investment in agric. 28 84 52 Investment in other industries 4 - 32 Consumption 4 Others 8 100 100 100 - 149 - Housing, as in most of the other survey areas, has highest priority. Investment in agriculture comes second. Additional consumption - surprisingly - is not of any importance. Farm programming As activities for programming Ruvuma farms, the following crop or crop mixes are taken in both their average and best productivity: Maize Maize-millet Maize-beans Tobacco In addition, the fertilized maize-beans mixture and sorghum (average productivity) are considered. Constraints relate to the present average farm size of 3.24 ha, the average level of working capital (variable costs) of 572 shs anO a monthly supply of family labour of 187 hours. The optimum solution yields a farm income of 4,275 shs from the cultivation of: 1.23 ha Maize (present average productivity) 0.61 ha Maize-beans (fertilized) 0.07 ha Tobacco 1.33 ha remain unused and family labour supply is constraining in the months of January, May, June and December. For these months subsequent availability of hired labour has been assumed (100 hrs/month), resulting in a change in cropping pattern as follows: 1.96 Maize (present average productivity) 0.82 Maize-beans (fertilized) 0.21 Sorghum (average productiivty) The gross margin increases to 6,452 shs/farm. 0.25 ha farm land remains unused. Labour, in spite of an increase of 100 hours per month of hired labour is constraining during the month of January, June and December. Farm income has to be lowered to 6,162 shs to allow for the payment for hired labour. No further programming is attempted, because it is not at all certain that more than the assumed use of hired labour would be forthcoming in reality. In Ruvuma, it seems that of all factors, labour is the most constraining. Assuming moderate levels of hired labour use during critical months, farm income of the average farm can be markedly increased. -150 - 2.1.10 Cereal Farming in Singida The Singida area, which is delineated by the divisional census areas 1412, 1416, 1431-2, and 1437, is characterised by typical dryland cereal farming, mixed with pulses, root and oil crops. 67 farmers were interviewed and the information they provided is almost complete. Missing is the specification of variable costs, such as, for seasonally hired labour, seed and other material inputs, as well as the value of farm- produced home consumption. All other information is summarized in Table 2.64. The Singinda farmer is 49 years old and has 0.6 years of formal schooling. He is head of a farm family of 6.2 people. The man-equivalents average 4.1 per farm and seem to increase with the increase in farm size and farm production. Farm sizes average 4.35 ha. This size is held by the majority of farmers; only the group with the highest level of farm production hold distinctly larger farm sizes. Crop production in Singida is dominated by a mixture of millet, maize and sorghum which occupies 73 percent of total cultivated area. 7.7 percent is occupied by the same crops butin purestand, while another 17 percent is sown to pulses, root or oil crops, and the rest to a variety of other minor crops. An average 4f 5.7 head of cattle are kept. They are predominant in the farm group with highest farm production. Labour inputs average 2,460 hrs per farm half spent on crops and half on livestock production. The value of farm production averages 2,772 shs. It is almost double this on the best farm group. Crops seem to contribute to it according to their proportion in the cropping pattern. The picture is, however, distorted by the inclusion of livestock which yield 35 percent of total farm production (ranging from 10 to 45 percent) but are mainly found on the best performance farms. Variable costs are not available and the value of farm production is taken as farm income. To it, off-farm income of 350 shs is added, which is highest in the best farm group. Total or family income amounts to 3,122 shs. Deducting household expenditures, a balance of 1,347 shs is left but this includes the value of farm produced home- consumption and variable costs. Comparing this balance with figures from other survey areas, it is doubtful whether any farm group,except the best, attain any positive surplus which could be used for investment or to meet other obligations. - 151 - STMARY TABLE 2.64 Composite Characteristics of Singida Farms - grouped by level of farm production Level of farm production VA Details iiAverage, Very Ve ry i I I high High Medium Low low No. of observations 13 13 13 13 13 6 Farner, family, labcLr Farmer's age, years 5005 50.9 44.1 45.5 53.2 48.8 Farmer's education, school yrs 0.5 - 0.5 0.9 1.1 Farm family, people 7..8 6.4 5.2 6.2 Perm. hired labour, nos. - 5 - Total people on farm 7.8 5.8 6.4 5.2 5.5 6.2 Labour availability, man-equiv. 5.1 4.3 4.2 3.5 307 4.1 FARM LAND Farm size, ha . 6.69 4.43 3.67 3.49 3.44 4.35 (standard error) (0.67) (0.57) (0.41 (0.49) (0.43) (0.27) Cropping pattern, % Millet-Maize-Sorghum 80.8 1 71.9 76.1 77.1 48.2 72.8 Maize-Millet-Groundnuts 6.0 8.1 5.2 6.5 15.5 7.7 Pulses, roots 11.8 20.0 18.7 10.1 28.9 16.8 Others 114 - - 6.3 7.4 2.7 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 Livestock, head of cattle 19.6 5.1 3.0 06 0.3 5.7 Labour inputs, hrs 4,393 2,911 2,479 1,674 1,082 2,460 I Percentage distribution Crop production 46.5 34.0 37.7 53.0 74.0 45.2 Livestock production 3.5 66.0 62.3 47.0 26.0 4 TOTAL 10.0 10I 100.0 100.0 100.0 100.0 __ __ __ __ __ __ __ __ _1_ 1000.00. Value of farm production, shs 5,330 3,476 2,389 1,741 919 2,772 (estandard error) (381.4) (81.0) (100.0) (34.6) (94.2) 207.3) Percentage distribution: Millet-Maize-Sorghum 45.9 31.9 46.1 65.1 54.0 45.3 Maize-Millet-Groundnuts 2.0 3.3 5.8 2.9 6.1 3.4 Pulses, roots 9.5 19.7 18.8 11.0 28.2 15.2 Other crops 0.6 - - 4.2 1.6 0.9 Livestock 42.0 45.1 29.3 16:8 10.1 35.2 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 /o - 152 - SUMARY TABLE 2.64 (cont'd) Level of farm 2roduction Detail .Average Very Very high High Medium Low low Value of farm production, she 5,330 3,476 2,389 1,741 919 2,772 Sum of gross margin, shs (farm income) L/ 5,330 3,476 2,389 1,741 919 2,772 Off-farm income, shs 1,200 - - 530 - 350 Family income, shs 1/ 6,530 3,476 2,389 2,271 919 3,122 Value of home consumption, shs n.a n.a n.a n.a n.a n.a Household expenditures, shs 2,617 1,624 1,484 1,580 1,570 1,775 Cash for investment, replacement, savings, 3,913 1,852 905 691 -651 1,347 consumption j/ LAND PRODUCTIVITY (crops) Value of production, shs/ha 466 433 495 420 235 419 Variable costs, shs/ha - - Gross margin, shs/ha - - - - Labour input, hrs/ha 305 226 255 254 232 260 LABOUR PRODUCTIVITY Value added, shs/man-equiv. 1,045 808 569 497 248 676 Value added, shs/labour hr. -used 1.21 1.19 0.96 1.04 0.85 1.10 Ratios Value of output to value of input not applicable a. Labour not costed b., Labour costed Degree of family labour use,%, 57 45 39 32 19 41 Degree of commercialisation, not applicablei Farm income as percent of total income 1/ 82 100 100 077 10 89 Iash balance as percent of total income 1/ 60 53 38 30 negative 43 Variable costs not accounted for 2j Variable costs and value of home consumption not regarded. - 153 - Farm sizes average 4.35 hectares and areimuch larger than in the other survey areas. Even the smallest farm is larger than one hectare. As can be seen below, farm sizes are rather equally distributed among farmers, in spite of the existence of a number of farms of more than average size. Farm size group, ha Average farm Percent of Percent of size, ha farms area 1.00 - 1.99 1.63 10.8 4.1 2.00 - 2.99 2.51 21.5 12.6 3.00 - 3.99 3.54 23.1 18.9 4.00 - 4.99 4.61 12.3 13.1 5.00 - 5.99 5.33 10.8 13.1 6.00 and over 7.67 21.5 38.2 TOTAL 4.35 100.0 100.0 The cropping pattern is shown in the table below. Crop or croD mix Ha Percentage on total Millet-Maize-Sorghum 2.96 69.0 Maize-Millet-Groundnuts 0.53 12.4 Millet 0.17 4.0 Millet-Sorghum-Cotton 0.16 3.7 Bananas-Cassava, beans 0.10 2.3 Maize-Sweetpotatoes, cassava 0.08 1.9 Maize 0.07 1.6 Groundnuts 0.05 1.2 Sorghum 0.04 0.8 Others (cashew nuts, rice, beans., sweetpotatoes, cassava, cotton, rice-cassava-bananas, bananas) 0.13 3.0 4.29 100.0 A mixture of millet and maize and/or sorghum covers more than 2/3 of the total cultivated area. 12 percent is under a mixture of maize-beans and/or groundnuts, while the rest is taken up by all other crops or crop mixes. - 154 - Mil let-Maize-S orghum Data are available from 88 farmers growing a mixture of the following (in almost equal proportions): Crop mixes % of the cultivated size of this mixture Millet-maiie-sorghum 25.8 Maize-sorghum 24.7 Milletwmaize 21.5 Millet-sorghum. 19.4 Millet-maize-others (sweetpotatoes, cassava, 8.6 groundnut s) TOTAL 100.0 The average field size under the millet/maize/sorghum mixture is 2.14 hectares, varying from 0.26 to 6.4 h40tares. The cultivation size 7per farm, because farmers may be cultivating more than 1 field, is 2.96 hectares. The average cultivation size of those farmers owning work-oxen is increased to 6.20 hectares, more than double the average. Unfortunately, no other inputs than labour are recorded. This does not permit a full enterprise analysis to be carried out; only labour inputs and cultivation size are available in the analysis as independent variables. On average 246 labour hours are used to cultivate onehectare of this crop mix, with relatively little variation. 28 percent of the total labour input is used for land preparation and planting, 37 percent for weeding and the remaining 35 percent for harvesting. Yields of each component are not known, but the compound value of production from this crop mix amounts to 410 shs/ha or 1,188 shs/farm. As material inputs are not recorded, the gross margin cannot be calculated. Equally, no expenses for seasonally hired labour are available. Assuming that variable costs are very small, the gross margin would roughly correspond to the value of production. The input/output relationships for this crop nix, stratified according to different productivity levels, are shown in Table 2.65. Values of production range from 196 to 901 over the farm groups. Highest productivity is achieved on smaller acreages than average and with higher labour inputs, particularly in weed control. TABLE 2.65: Input/Output Coefficients by Level of Productivity - Millet/maize/sorghum, Singida - Input/output coefficients _________Level of productivity Average Very I i Very high High Mdium Low low No. of observations 17 17 1 18 18 88 Cultivation size, ha 1.37 2.06 2.20 2.45 2.57 2.14 Labour inputs, hrs/ha Land preparation, planting 97 40 60 49 40 55 Weed control 181 95 106 76 75 103 Harvesting, others 143 984 65 63 88 TOTAL 421 234 250 190 178 246 Vaife of production, shs/ha 901 431 395 301 196 410 It is therefore not surprising that Cobb-Douglas production function yielded a high elasticity for weeding labour (0.54*), and also for land (0.29). Labour for land preparation yields only 0.04. The sum of elasticities amounts to 0.87, with an adjusted R2 of 0.54- Marginal productivity for land is 86 percent of assumed cost of 138 shs/ha; for labour in land preparation it amounts to 30 percent, and for labour in weed control to twice the factor costs. Using the above results to find the optimum resource combination, farm sizes should be decreased to 1.58 ha; labour in land preparation becomes negligible and should be used in the least amounts necessary, while labour in weed control could be profitably increased up to 242 hrs/hao Returns per ha could thus be increased to 637 shs, which is 55 percent more than presently achieved on average. - 151 - Maize-Millet-Groundnuts 23 observations are available on this ty'pe of mixture, which is made up of maize and millet and/or groundnuts, beans or cassava. Their percentage distribution is as follows: Crop mix Ha Percentage on total cultivation size Maize-millet-groundnuts 0.68 45 Maize-(millet)-beans 0.32 22 Maize-(millet)-cassava 0.42 33 1.49 100 The average field size is 1.49 hectares and the cultivation size of this mixture, on average fDr all farmers in the survey is 0.53 hectares. Labour inputs total 279 hrs/ha and are very similar to those of the previous mix. 30 percent of total labour input was used for land preparation, 34 percent for weeding and 36 percent for harvesting. The composite value of production of all components of the mix amounts to shs 441/ha, which is also comparable with the first mix. Production function analysis yielded positive elasticities for land (0.52) and labour in weeding (0.27). Labour in land preparation had a negative elasticity. The sum amounts to 0.73 with R2 of 0.44. Marginal nroductivity for land is 66 percent higher than assumed factor costs and for labour in weed control it is 27 percent higher. An optimal factor combination is very similar to present averages (0.65 ha and 90 hrs/ha in weed control). No higher than average returns are calci Ked. Table 2.66 shows the input/output relationships for this mixture, stratified by level of productivity. TABLE 2.66: Input/Output Coefficients by Level of Productivity - Maize-millet-groundnuts - Singida - Input/output coefficients Level of productivity Average Very IVery high High 1 Medium i _-Low low No. of observations 4 4 5 5 5 23 Cultivation size, ha 1.11 0.93 1.85 _2.00 1.46 1.49 1Labour inputs, hrs/ha Land preparation, planting 85 120 76 85 Weed control 171 137 63 1 9 43 94 Harvesting, others 178 229 6 63 100 TOTAL 434 486 211 268 182 279 Value of production, shs/ha 847 611 : 456 363 196 441 1 - 150- Other crop and crop.mixes As has been noticed before, a number of other crops or crop mixes are grown, all of which are, however, cultivated on a limitedscale and account all together for less than 20 percent of the total cultivated farm sizes. Grown in purestand are millet, sorghum, maize, cassava, sweetpotatoe and groundnuts. In addition the following mixes are observed: Banana-cassava-beans Millet-sorghum-cotton Maize-cassava-sweetpotatoes Rice-sweetpotatoes-bananas The input/output coefficients of these crops and crop mixes are summarized in Table 26.7. Livestock Nine farmers, or 13 percent, own on average 3 work oxen, valued at shs 813 each. No records are available of their actual utilization in crop cultivation. Another 5 farmers own donkeys, 5 each, valued at shs 448/head. Productive livestock in the form of cattle is kept by 28 farmers or 42 percent of all farmers. An average 13 head are kept, valued at shs 243 per head. Of these, 4 were males and 9 cows and/or heifers. Their distribution is rather uneven: while the large majority of cattle keepers held an average of less than 10 head, 3 farmers alone held around 50 head each. All cattle keepers, except 3, also keep sheep and/or goats. Sheep are kept by 28 farmers with an average of 11.4 head, valued at 39 shs/head. An overall average for Singida farms of the number of sheep kept is 4.6 40 farmers keep goats and the number kept averages 14.3 (range: 1-97). Goats are valued at 42hs/head. Over..all farmors the number of goats kept averages 8.8. Labour inputs per livestock keeping farmer amount to 3,190 hours and - except for the farms with large herds - are not dependent on the number of aniials/farm. Livestock has to be herded whether the herd consists of 2 or 40 head. The value of production consisting of both consumption and -ale of livestock products and inventory changes, amounts to 2,412 shs per livestock keeper, ranging from 300 to 6,862 shs, the latter being achieved by farmers with largest herds. On average over TABLE 2.67: Input/Output Coefficients by Luvel of Productivity - Minor crops, Singida - Maize/ .Sweet oRice/ Input/output !Banana/ 'Millet/ Cassv/ Potatoe Sm aizo Cassava S.9otato/-Grcund- coefficients Cassava/ Millet o S.Pota.t Banan nut coeficintsBoans :Cotton SPtt aa u No. of observations 9 8 6 4 4 4 - 3 2 2 2 Cultivation sizo, ha 0.71 1.. 174 1.2 0.41 0.73 1.63 0.8 0.8i- 1.58 Yield, kg/ha - 91 1,191. 172 94 459 - 55 Value of production shs/ha 8822263 326 l13307 715 31445152 2292537 327 I Labour inputs, hrs/ha ' Land preparation, planting I323 86 92 183 461 112 38 165 315 55 Weed control 121 69 148 290 81 140 61 - 79 22 Harvesting, others ' 206 41 145 192 213 90 29 99 123 62 -TOTAL 550 196 385 665 755 342 128 264 523 139 Distribution by season Jan 86 34 99 29 218 125 38 121 252 51 Feb 172 14 - 220 173 : 24 15 22 12 4 Mar 77 16 57 45 70, 40 8 I45 29 - Apr 139 - - - 227 14 - 76 82 22 May 51 31 - 11 45 26 - - - - Jun . 10 79 13 - . 45 29 - 95 Jul 3 - 67 212 - - - , 2 Aug 3 - - Sep 3 23 - 37 - Oct 3 40 12 68 31 - - 40 62 12 Nov 3 62 15 - .Dec 3 28 21 71 21 11 7- 53- TOTAL 550 196 385 665 755 342 128 264 523 139 all farms, returns from livestock amount to 976 shs or 35.2 percent of the total value of farm production. This indicates the considerable importance of livestock in the farm organization of Singid. farms. Whole farm With an average cultivated farm size of 4.35 ha and a livestock herd consisting of 5.7 head of cattle and 13.5 sheep/goats, thu Singida farmer earns a value of production of 2,772 shs. The labour input averages 2,508 hrs. The amount of variable inputs is not known but is considered low. Better farms are larger both in crop area, number of animals kept and amount of labour used. Productivity of land is, however, comparable for all farm groups except the lowest. Higher farm production is thus more than anything else a function of the resource base in the form of leand and livestock. (Table 2168) Cobb-Douglas function analysis yielded pesitive elasticities for all variables (land 0.29*; livestock 0.15; labour 0.58*; sun 1.02; R2 0.65). Marginal productivity is only higher than its assumed factor costs (138 shs/ha) in the case of land, supporting the conclusion above that farm production is increased through expansion in farm size, if at all possible. Optimal size is indicated to be 9.2 ha. yielding 3,600 shs at a productivity rate lower than at present. In Table 2.69, the income position of the Singida farmer is assessed. Since variable costs are not available, the value of farm production has to serve as proxy for income. It averages 2,722 shs/farm, ranging from 919 to 5,330 ovor the farm groups. Off-farm income is earned by only a few farmers who already mainly belong to the farm group with highest production. This accounts for a wider disparity in family income ranging from 919 to 6,530 shs/farm. Household expenditures which average 1,775 shs, are uniformly high across farm groups, except in the best group where they are 47 perccnt higher than average. Major components of expenditure items are clothing and alcoholic drinks. A balance amounting to 1,347 shs is left which contains the (unspecified) value of farm produced home consumption, variable costs and possible cash surplus (or deficit)for investment, replacement or savings. For the farm group with lowest production a negative value is already evident, even before accounting for above. For the other farm groups, the balance is so marginal that if values for farm produced home consumption and variable costs of similar survey arecs were inputed, it is very doubtful whether a breakeven point could be reached. - 16f) - Farmers' attitudes In Singida area information on farmers' a titudes is not available, except for their preference, given additional incomes: Preference, percent 1st 2nd 3rd Construct house 50 27 17 Invest. in agriculture 48 70 80 Invest. in other industr. 2 3 3 To construct new and better houses had highest priority in the farmers' minds. This satisfied, investment in agriculture is indicated as gradually increasing. Farm proganming For LP both the two most important crop mixes mentioned before viz: millet-maize- sorghum and maize-millet-groundnuts, in both their average and best productivity are considered. In addition, livestock is entered. Constraints relate to the present average farm size of 4.35 ha, present number of cattle (5.7 head) and the family Iabanr supply of 512 hrs/month. The level of working capital is not available. As expected, the high level of family labour supply allows the whole available farm- land to be cultivated. Labour becomes constraining in May. Present average herds of 5.7 head of cattle can be kept in addition to the cultivation of both crop mixes in their best productivity in the following proportions: Millet-rmaize-sorghun 2.95 Maize-millet-groundnuts 1.40 This is not different from present practices, except that best productivity levels should be the aim. As a result, the value of production could be raised to 4,818 shs, which is 74 percent higher than the present average and only slightly lower than the level achieved presently by farmers with highest production, although on much larger acreages. - 16* - TABLE 2.68 (a) Input/Output Coefficients for Farms by Level of Farm Production - Whole farm, Singida - Level of farm production Input/Output coefficients I Average Very Very high High Medium Low low No. of bbservations 13 13 13 13 13 65 Cultivation size, ha 6.69 4.43 3.67 3.49 3.44 4.35 No. of livestock 19.6 5.1 3.0 0.6 0.3,. 5.7 Labour input, hrs. 4,393 2,911 2,479 1,674 1,082 2,460 Value of farm production, 3,476 1,741 919 2,772 5,330 346 2,389 .il 1 t7 (b) Input/06tput Coefficients for Farms Grouped by Farm Size - Whole farm, Singida - Farm size group, ha Input/Output coefficients 1.00 - 2.00 - 3.00 - 4-00 - 5.00 - 6.00 & 1.99 2.99 3.99 4.99 5.99 over No. of observations 7 14 15 8 7 14 Cultivation size, ha 1.63 2.51 3.54 4.61 5.33 7.67 !No. of livestock 5.0 1.2 3.0 9.2 - 10.7 ILabour input, hrs 2,261 1,667 2,510 3,050 1,491 3,327 ,Value of farm production 1,590 2,001 2,644 3,009 2,171 4,388 10,1. - 16).- TABLE 2.69 (a) Farm End Family Income and Household Expenditures - Singida - - shs - Level of farm production Details Averake Very Very high High Mediun Low low Value of-production Farm income1/ 5,330 3,476 2,389 1,791 919 2,772 Off-farm income 1,200 - - 530 - 350 Family income 6,530 3,476 2,389 2,271 919 3,122 Household expenditures 20617 1,624 1,484 1,580 1,570 1,775 Balance Value of home consumption, cash for investment, 3,913 1,851 905 691 -651 1,347 replacement, savings (b) Composition of Household Expenditures Item Shs Percentage of total Clothing 707 39.8 Alc. drinks 480 27.1 Kerosine 158 8.9 Cooking oil 146 8.2 Condiments 132 7.4 Detergents 118 6.7 Others 34 1.9 TOTAL 1,775 100.0 l1/ Variable costs not accounted for. APPENDIX A SURVEY AREAS WITH PARTIAL INFORMATION Due to a number of reasons only partial information is available from a number of survey areas. These are: Survey area Divisional census areas Westlake 1731 - 4 Arusha 0122 - 3, 0144 Kondoa 0331, 0334, 0337-8, 0321 - 8. Tanga 1621, 1631, 1633 - 5, 1637, 1641, 1643, 1662, 1665 - 6. Mtwara 1051, 1053 - 4, 1091, 1095 Mwanza 1111, 1116 - 7. Mara 0723, 0731, 0736 - 7, 0739. Because of the general lack of farm management information in the country, rather incomplete data thought to be useful have been compiled and analysed as far as possible in the following pages. - 16#'- A.1. Coffee-Banana Farming in Westlake Westlake area is defined by the divisional census area 1731-4. Some information is available from 43 farmers and is summarised in Table A.1.1. The main handicap to analysis is the fact that only cash values were recorded, while farm produced home consumption as well as inputs were omitted. The Westlake survey area was previously investigated in 1964/65 and results of both surveys will be compared whenever possible. Generally, the earlier study contains much more detailed insight into the farm organization than was possible to collect in the more recent effzrt. The Westlake farmer is 54 years old, has two years of schooling and headsa farm family of 5.8 people (5.3 who form a work force of 2.9 man-equivalents. The Westlake cultivated farm size averages 0.78 ha (0.88 ha), of which 79 percent is under coffee and bananas (89 percent). The best farm group cultivated an average of 1.42 ha (1.49). Farms in this group have a bigger resource base and have, in addition to more land, larger families and work forces and more livestock. Over all farms,an average f 2.0 head of cattle are kept (1.6 head). Cattle are concentrated and in the best farm group, averaging 6.3 head (5.3) per farm. Labour inputs per farm are much under recorded for reasons explained earlier and di not render themselves for comparison. The cash value of production aicunts to 838 shs per farm. Assuming that this value is derived from coffee and 25 percent from other crops, it not only corresponds to the earlier survey results (834 shs), but also indicates that the 'stagnation at a higher level' described in the earlier study has continued and little progress, despite the suggested ways of development, has been nade. 1,/ K.H. Friedrich, "Coffee-banana holdings at Bukola" - in : Ruthenberg, H. Smallholder Farming and Smallh:lder Development in Tanzania, 1968. 2./ Figures in parenthesis refer to the results.of earlier surveys. SUMMARY TABLE A.1.1 Composite Characteristics of Westlake Farms - grouped by level of farm prochictinn cash - Level of farm production Dletails VrVey Average Very Very high High Medium Low low No. of observations 9 9 9 8 8 43 Farmer, family, labour Farmers' age, years 48.3 58.0 59.7 48.4 56.1 54.2 Farmers' education, schoo.l-yrs 3.2 3.3 1.4 1.9 0. 2.1 Farm family, people 8.7 5.9 4.9 4.1 5.0 5.8 Perm. hired labour, nos. 0.1 - - - Total people on farm 8.7 5-9 5.0 4.1 5.0 5.8 Labour availability, man-equiv 3.7 2.9 2.9 2.3 2.4 2.9 FARM LAND Farm size, ha 1.2 0.89 0.81 o42 0.28 0.78 (standard error) (0.23) (0.17) (0.19) (0.07) (0.07) (0.09) Cropping pattern, % coffee/bananas 77.5 80.6 73.2 88.0 1 85.7 79.5 Roots, pulses 19.2 14.3 23.2 8.0 10.7 17.0 Cereals 3.3 5.1 1:6 4.0 3.6.. 3.5 TOTAL 100.0 100.0 100.0 :100.0 100.0 100.0 Livestock, head of cattle 6.3 0.7 2.0 - 0.6 2.0 Variable inputs, shs 1/ 77 21 17 2 10 26 Percentage of distributi.n: Fertilizer - 52,0 88.0 - 70.0 26.9 Pesticides 27.0 - - 16.9 Hired labour 73.0 48.02.0 100.0_30.0556.2 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 Labour inputs, hrs 3,621 937 1,017 436 321 1,308 Percentage distribution: Group production 57.3 28.1 32.6 : 67.0 54.5 Livestock production 42.7 71.9 67.4 33.0 45.5 0.6 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 Value of farm production, shs 1/ 2,445 957 481 128 i 838 (standard error) 407.7) (61.5) (33.2) ,(26.0) (:.) (160.4) I/ Cash only - 166- SUMMARY TABLE A.1.1 (cont'd) Level of farm nroduction Detail Ver Vey Average Very IVery high High Medium Low low Value of farm production shs 1/ 2,445 957 481 128 7 838 Variable costs, shs 77 21 17 2 10 26 Sum of gross margins, shs 2,368 936 464 126 -3 812 (farm income) Off-farm income, shs 2,022 547 207 381 723 786 Family incomc, shs 1/ 4,390 1,483 671 507 720 1,598 Houschold expenditures, shs 2,516 2,684 1,447 1,232 670 1,745 Cash for investment, replacement, savings, 1,874 -1,201 -776 1 -725 50 -147 consumption Cash only Parms with highest cash production achieved 2,445 shs (1,967 shs) which is three times higher (2.5 times) than average. For both surveys this is achieved through increase in scale of farm operations, but also through increase in productivity which is 60 percent higher than average (40 percent). 28 percent of farmers have off-farm employment and, on average, 108 days per year at 15.70 shs per cay, thus earning 786 shs. Farm sizes average 0.78 ha. varying from 0.38 to 1.,42 over all farm groups. It can be seen from the table bel-w that 70 percent of all farmers have holings smaller than 1 ha and cultivate less than 38 percent of the total available farm land. - 16 - s zc groups. ha Average farm Percent of Percent of total size, ha farmers area 00 0.25 39.6 12.6 - 0.09 0.65 30.2 25.2 - . 149 1.18 11.6 17.6 13 - 9 1.76 14.0 31.4 C 2 _9 2.22 4.6 13.2 L 5D uin".er- 0.78 100.0 100.0 'he croppin] pattern is dominated by a mixture of coffee and banana. Coffee in p~resto.d accounts for less than 10 percent of the total area under perennials, which cover 79,5 percent of the total cultivated area. Root crops and pulses are grown on 17 p,rcent and cereals on 3.5 percent of the cultivated area. There seems to be no definite trend in the cropping pattern over the farm groups. Coffoe/brna ,*9 observations relate to this crop mix, whose input/output coefficients are s-m=nrized in Table A.1.2. Cultivation size averages 051 hectares. Analysis is distcrted by having to rely on cash production only. The value of h ome consumption -,ris not recorded. Nevertheless, Table A.1.2 shows that there are farmers who make three timea 1he average cash value of production. They achieve this on slightly smaller than average acreages, having more banana trees in the mixture, applying both fertilizer and pesticides and using more labour in crop cultivation. Cobb-Douglas production function analysis yielded the following elasticities: land 0.15; fertilizer 0.01; pesticides 0.23; no. of banana trees 0.33*; no. of coffee troes 0.26*; labour in cultivation 0.22. The sun of elasticities is 0.74; the adjzsted -2 is 0.47. Marginal costs are higher than assumed factor costs only for fertilizer. The optimal resource combination would be: 0.38 ha 26 shs fertilizer costs, 211 banana trees,317 coffee trees and 939 labour hours, yielding a cash value of production of 1,049 shs or a productivity of 2,760 shs/ha, which tends to confirm the above findings, TABLE A.1.2 Input/Output Coefficients by Level of Productivity -Coffee/banana, Westlake - Level of nroductivity Input/Output Coefficients Very Very Average high High Modium Low low No. of observations 9 10 10 10 10 49 Cultivation size, ha 0.42 0.41 0.58 0.65 0.46 0.51 Variable inputs, shs/ha 1/ Fertilizer 35 - - - - 5 Pesticides 15 - l0 - 24 9 Hired labour - 2 11 10 2 6 No. of bananas 764 495 263 318 108 320 No. of coffee trees 486 418 316 529 442 481 Labour inputs, hrs/ha Weed control 441 419 222 207 332 285 Harvesting, others 243 16 104 74 79 120 TOTAL 684 588 326 281 401 405 Value of production shs/ha /26626 1,074 659 393 149 862 / Cash only Coffee Eight obsbrvations are available on purestand coffee prouction. Cultivation sizes average 0.36 and the number of coffee trees 625 per ha. Labour inputs amount to 544 hrs/ha, which is higher than for the crop mix coffee/bananas. Yields average 455 kg valued at 1,365 shs/ha. Best performance is observed on smaller cultivation sizes with a higher number of coffee trees/ha. and much higher labour inputs. Productivity is raised by 50 percent compared to the average (Table A.1.3). Cobb- Douglas production function analysis yielded the following production elasticities: land 0.30; labour 0.17; coffee trees 0.60. The sum of elasticities is 1.07 and the adjusted K2 059. Optimum resource combination requires cultivation sizes of 0.15 ha only. Labour per ha would have to be increased to 553 and number of trees/ha to 1,953. This combination would yild a productivity of 3,317 shs/ha which is 'J times higher than the pnesent average and 60 percent above present best performance. TABLE A.1.3 Input/Output Coefficients by Level of Productivity - Coffee, Westlake - Input/output coefficinets iLevel of pro6uctivity Average High Low No. of observations 4 - 4 8 Cultivation size, ha 0.31 0.39 0.36 Variable inputs, shs/ha Hired labour 412 31 No. of coffee trees 698 565 625 Labour inputs. hrs/ha Land preparation, planting - Weed control 457 290 372 Harvesting, others 230 119 .2 TOTAL 687 409 544 Yield, kg/ha 683 268 455 Value of production, shs/ha 2,055 804 1,365 Other crops - A variety of other crops is grown apart from coffee and bananas. These include sweet potatDes, cassava, groundnuts, babbaranuts, millet, rice, maize and tea. For all of them no complete records are available, especially with regard to yield, variable costs and labour in harvesting. Cultivated areas for all of these crops are very small. The information on these crops, which is available, is summarized in Table A.l.4. Livestock 13 farmers,or 30 percezt of all farmers, keep cattle. The number of animals kept averages 7, ranging from 1 to 18 head, and valued at 641 shs/head. 9 of the cattle keepers also have goats. Goatr are held by 15 farmers, who koep 3.4 animals on average, varying frcm 1 to 12, each being valued at 71 shs. - 170 - Five farmers keep poultry, 16 birds on average (3-21) and valued at 14 shs/bird. The cash value of production amounts to 162 shs on average of all farms. This magnitude, h,wever, the value of h me consumption being unrecorded, does not allow the true productivity of livestock keeping to be estimated. Labour inputs f-r livestock average 662 hours per farm in Westlake. Whole farm The cash value of farm production in Westlake area averages 838 shs. The value of farm produced hnme consumption is not available as are farm produced inputs (such as seed) and labour inputs into certain operations (e.g. harvesting). It is for this reason that farm productivity cannot be computed and comparison is made impossible. Neverthe- less, one can say that farms with highest cash values make almost three times more than average on nearly double average acreage. Both cash costs, labour inputs, and number of animals kept are substantially higher than in any other farm group. (Table A.1.5) Grouped by farm size, the value of cash production seems very much to be depen,'ent on both scale of operation expressed in hectares and number of cattle kept, while the latter in turn seem to determine the magnitude of the farm's labour input. The farm cash income averages 812 shs per farm. Off-farm income accrues to all farm groups, but is in the best farm groups more than twice the average. This fact influences the disparity in total family cash income: although incomes average 1,598 shs, they range from a low of 507 shs to a high of as much as 4,390 shs/farm. (Table A.1.5) Household expenditures, in contrast to many of the other areas, have a declining tendency from 2,500 shs in the highest productivity farm group to 670 she in the lowest. They average 1,745 shs, being spent on avriety of items of which clothing, sugar and flour are the most important. The cash balance is negative (-147 shs). Only for farms with the highest cash production and coinciding with higher levels mf off-farm income is there a positive balance averaging 1,874 shs per farm, which represents 42 percent of total cash income. Farmers' attitudes Only some indicative information is available about farmers' attitudes and reaction to changes: producer prices especially for millet and bananas are regarded as too low. Coffee is not mentioned. TABLE A.1.4: Input/Output Coefficients by Level of Productivity Input/output Sweet- Cassava !Groundnut Millet Millet/ Rice Maize Ground- Bambara- Tea coefficients Potatoe Cassava Cassava nut nut No. of observations, 2 2 10 7 5- 2 3 Cultivation sizes, ha 0.12 0.11 0.19 0.09 0.13 0.10 0.06 0.06 0.06 0.31 Yield, kg/ha n.a n.a 1 345V- 8oo 333 n.a. 382 10104 Value of production, 2 2 a8 shs/ha n.a. 454- n.a. 2,172- 983 2,000 333 n.a. 1,145 740 Labour inputs, hrs/ha Land preparation, planting 622 476 543 633 530 1,364 550 542 953 - Weed control 137 i111 199 34 - 392 142 200 558 85 Harvesting, others n.a n.a n.a 392 i455 777 8 n.a 264 203 TOTAL n.a n.a n.a : 1,059 1 985 2,533 700 n.a 1,77, 288 Jan 194. 71 158 380 78 487 - 113 421 37 Feb 124 101 51 6 - 1 208 133 42 .20 . MAr 72 60 8 128 - 300 25 83 253 142 Apr 121 29 1231 16 95 451 - 121 42 20 tr50y 51 59 '188 454 195 4 75 163 - Jun 46 23 39' 198 - 131 - 8 - Jul 10 - : - - - - - 421 47 Aug 9 :8 - .:2 - 469 333 - 37 - Sep2 - - 259 25 - Oct 17- 38 - - 211 - Nov 23 1 - 21 - - . 162 67 - 184 - D00 1021 191 314 31 358 1 - - 1208 - TOTAL n.a1/; n.a1/ n.a 1/1 1,059 985 12,533 1 700 n.a/ 1,774 288 1/ Harvesting not included 2 4 observati-ns only. - 17 L - With regard to inputs, 20 percent of the farmers complain about the unavailability f cow mcanre and its high cost. The same is true to a lesser extent for mulch grass. 30 percent of all farmers hold the same opinion about fertilizer and pesticides. Only insufficient insight could be gained with regard to farm re--rganization should existing restraints be -removed. It was only evident that almost half of all farmers would like to keep cattle and in much greater numbers than at present. Finally, farmers preferences were tested with regard to the use of the eventually higher farm incomes. Their responses were as follows: Preference, percent 1st 2nd 3rd Build house 44 17 13 Invest. in agric. 28 63 9 Invest, in other industries 4 - A Consumption 22 12 61 Others (marriage, etc) 2 8 13 Building of a new house has first priority, followed by investment in agriculture. After these needs are satisfied, additional incomes would mainly be spent on aeditional consumption. I/ TABLE A.1. (a) Input/Output Coefficients by Level of Cash Production - Whole faxm, Westlake - Level of cash production J'Input/Output Coefficients Very Averaga VerY Vr high High Medium Low low No. of Dbservations 9 9 9 8 8 42 Farm size, ha 1.42 0.89 0.81 0.42 0.28 0.78 Head of cattle 6.3 0.7 2.0 - 0.6 2.0 Labour in-_uts, hrs 3,621 937 1,017 '36 321 1,308 Value of production, shs 2,445 957 481 128 7 838 Variable costs, shs 77 21 17 2 10 26 Sum of gross margins, shs 2,368 936 464 126 -3 812 (b) Input/output Coefficients by Farm Size - Whole farm, Westlake - Farm size grou-p) ha In put/output coefficientsUeFamsz ru,h- 0.50- 1.00- 1.50- 2.00- 2.50 & 0.50 0.99 1.49 1.99 2.49 over No. of observations 17 13 5 6 2 - Farm size, ha 0.25 0.65 1.18 1.76 2.22 - No. of cattle 0.9 0.5 5.8 2.7 9.5 - Tabour inputs, hrs 502 1,521 2,534 1,388 3,463 Value :f cash production, shs 370 557 1,668 1,358 3,010 - Cash costs, shs 12 22 11 17 241 Gross margin, shs 1/ 358 535 1,657 1,341 2,769 - 1/ Cash - 17# - TABLE A.1.6 (a) Farm, Family Cash Income and Household Expenditures - Westlake - -she- Details of income Level of cash -:roduction 4t. Average and expenditure Very Very high High Medium Low low Farm cash income 2,368 936 164 126 -3 812 Off-farm income 2,022 547 207 381 723 786 Family cash income 4,390 1,483 671 507 720 1,598 Family cash income 4,390 1,483 1,447 1,232 670 1,7,5 Cash for investment, replacement, savings, 1,874 -1,201 -776 -725 50 -147 consumption (b) Composition of Household Expenditures Item Shs Percentage of total Clothing 321 18.4 Sagar 320 18.3 Flour 232 13.3 Condiments 160 9.2 Detergents 160 9.2 Alcoholic drinks 148 8.5 Stimulants, of which 121 6.9 tobacco 103 5.9 Transport 88 1 5.0 Kerosine 73 4*2 Medicine 69 3.0 TOTAL 1,745 100.0 -17f'- SUPPLEMENT TO APPENDIX TABLE A..11 Labour ReEuirements Df Major Crops - Westlake - ftalHrs/ha Hrs sCoffee/banana C3ffee Farm Average Be st 1Average Best Distribution by season Jan r 2 111 278 99 Feb 5 - - - 72 Mar 62 83 82 71 3e6 Apr 99 184 100 60 136 May 75 195 85 111 129 Jun 74 68 57 106 116 Jul 26 27 32 76 72 Aug 39 68 13 25 84 Sep 10 17 41 103 64 Oct 3 23 - - 56 Nov - 3 3 8 57 Dec 7 14 19 19 97 TOTAL 105 684 524 687 1,308 Distribution by type of labour Family 1,00- 68.. 409 416 1,289 Seasonal hired labour - 135 271 19 TOTAL 405 684. 54 687 1,308 Distribution by operation Weeding, cultivating 285 '41 372 i57 Harvesting, others 120 243 172 230 TOTAL 05 684 i 54: 587 - 176 - A.2 Arusha Area Information from this area relates to 25 farms whose basic characteristics are summarized in Table A.2.1 TABLE A.2.1 Characteristics o)f Arusha Farms Detail j Farm size groups, ha Unt'er 0.50 - 1.00 - 1.50 - 2.co & 0.50 0.99_1.49 1.99 above No. of observations 2 8 2 3 3 2 Average farm size, ha 0.27 0.80 1.25 1.72 2.4 .00 % of farmers 36.0 32.0 8.0 12.0 12.0 100.0 total area 10.1 26.7 10.8 21.6 30.8 100.0 Farmer and farm family* Farmer's age 41.7 46.9 33.0 43.3 51.0 44.0 Narmer's education 1.8 0.7 - - 0.3 People/far 5.7 5.7 5.5 7.3 6.7 6.0 Man-equivalent/faru 2.7 3.0 2.3 3.8 4.5 3.1 Crojping pattern, ha Maize/beans/vegetables 014 0.20 0.61 0.53 045 0.29 1Joffee/banana 0.13 0.55 0.53 0.74 1.51 0.56 Household expenditures, she 2,761 2,782 1,817 2,419 3,315 2,718 68 percent of all farms are smaller than one hectare and cover 37 percent of the total farming area. Cultivated farm sizes averaCe one hectare. The Arusha farmer is 44 years of age; his farm household includes 5 people, in addition to himself, forming 3.1 man-equivalents available for far work. In the cropping pattern a mixture of coffee-bananas predominates with 52 percent ;f the total area. Both crops are grown also in purestand, but to a very limited extent (coffee 3 percent, bananas 1 percent). In addition to these crops a mixture of maize- beans is grown on 16 percent of the total area. In purestand the acreages 'f these crops cover 8 and 1 percent respectively. Vegetables are grown on 4 percent, and the rest 15 percent of the area is used as p:-x'e. - 17 - This cropping pattern changes with changing farm size. The percentage -f czffee/ banana mixture on the total cultivation size tends to increase with expansion in farm s-ze. HDusehld expenditures average 2,718 shs and are distributed as follows: TABLE A,2.2 Average Household_Expenditure by Item - Arusha - Expenditure item Average expenditures, shs Percent of total expendi.ture Flour, bread 824 30.3 Sugar 493 18.3 Cloth 390 14.4 Cooking oil 378 15.0 Alc. drinks 183 6.7 Stimulants: 144 5.3 Tea 71 2.6 Tobacco 60 2.2 Coffee 13 10.4 Rerosirne 116 ).2 Detergents 106 3.8 CondimentF 73 2.6 Others 11 0.:. TOTAL 2,718 100.0 Expenditure on flour and bread accounts for 30 percent of total expenditures. High also are expenditures on sugar and cooking oil. A.3 Kondoa Area i9 farmers provided basic farning statistics, summarized in Table A.3.1. below. 170 - TABLE A.3.1 Characteristics qf K-ndoa Farms Farm size ru Details e-goAv --erage nder 1.00 - 2.00- 3. 00 4.0 - 5.00 & 71.00 1.99 i 2.99 3.99 4.99 above No. of observations 3 13 ) 14 7 10 2 9 Cultivation size, ha 0.98 1 . 5 8 2.27 3.41 :.53 8.33 2.88 % of farmers 6.1 26.5 28.6 14.3 20.4 4.1 100.0 o of total area 2.1 14.6 22.5 16.9 1 32.1 11.8 100.0 Farmer and farm family Farmer'a age 45.7 36.2 1 46.8 44.1 39.1 44.0 41.9 Farmer's education - 0.6 - -- People/farm 3.0 55 5.7 o9. 5.6 Man-equivalent farm 2.2 2.9 3.1 3.7 3.3 5.7 3,4 Farm size averaees 2.88 ha, rather equally distributed among farmers, compared to other survey areas. Kandoa is mainly a cereal growing area. Maize, millet, sorghum (mixed with some beans and cassava) are groiwm in both purestand and different mixtures. Crop or crop mix. Cultivation size in Percent of total haZfarm cultivation size Maize/millet 0.84 29.2 Maize/sorghum/mil let 0.58 20.1 Maize 0.48 16.1 Millet (bullrush) 0.2: 8.3 Millet (finger) 0.15 5.2 Maize/rillet/cassava 0.14 4.9 Maize/sorehum 0.13 4.5 Sorghum 'Ir 0.10 3.5 Sorghum/millet 0.10 3.5 Others (maize/beans, millet/beans, cassava, sugarcane) 0.12 4.1 TOTAL 2.88 100.0 An attempt to differentiate the cropping pattern by farm size in order to determine the changes in crop mixes, as well as other anlysis, proved inconclusive. A, Tanpa Area The composite characteristics available from 39 farmers-in Tanga area are summarized in Table A.4.1. Farmers are 57 years of age and have a higher educational standard attained than farmers from many other areas. People per farm average 4.6, farming 3.0 man-equivalents. It is interesting, although only few observations are available# that larger farms are mainly managed by younger farmerst with more school years and larger farm families. Farm sizes average 1.24 ha of which half is in maize an' rice and half in a mix of cashewnuts, coconut, cassava. Smallest farms cultivate only cereals, while the proportion of perennial crops on the cropping pattern seems to increase as farms get larger. Equally household expenditures increase. TABLE A.4.1: Characteristics of Tanga Farms Farm size proup, ha Under 0,50 - 1.00 - 1.50 - 2.00 - 2.50 & Ave 0.50 0.59 j 1.49 1.99 2.49 over No. of observations 5 10 14. 8 1 1 39 Average farm size, ha 0.40 0.83 1.35 1.75 2.3 2.82 1 % of farmers 12.8 25.6 35.9 20.5 2.1 2.1 100.0 % of total area 4. 1 15 L2 391_ 28.9 4.9 5.8 100.0 Farmer & family Farmer's age 59.0 51.2 64.5 56.6 45.0 48 57.8 Farmer's education 1.2 1.2 2.8 2.4 5.0 10.0 2.3 People/farm I 3.6 4.9 4.3 4.9 5.0 8.0 4.6 Man-equivalents/farml 2.2 2.3 3.3 3.7 2.0 6.2 3.0 Croiinr -attern ha I Maize, rice, maize cassava 0.40 0.64 0.65 0.69 0.94 - 0.62 Ca-shewT/coconut/ II cassava - 0.13 0.70 1.06 - 2.82 0.57 TOTAL 1 0.40 0.77 1.35 1.75 0.94 2.82 1.19 Household expenditures 74 416 921 538 650 2,074627 - 18D- A*5 Mtwara Area 46 Mtwara farmers provided information on their farming situation which is summarized in Table A*5.1. TABLE A.5.1 Characteristics of 4twara Farms Unt F a r Farm size group,. ha Average Uhder 5 1.00- 1.50 - 2.00 - 12.50 & 0.50 _.99 1-49 l99 2.49 - over No. of observations 4 10 15 11 3 3 46 Cultivation size, ha 0.48 0.77 1.21 1.76 2.36 3 1.38 o of farms 8.6 21.7 32.7 24.0 6.5 6.5 100.0 o of tPtal area 2.9 11.9 23.2 30.0 11.0 16.0 100.0 Farmer, farm family Farmer's age 48.5 414 i42.9 49.1 550 38.0 46.0 Farmer's education - 0.4 0.0 0.7 - 2.7 0.7 People/farm 2.0 3.5 3.8 3.5 5.7 8.3 3.9 Man-eqaiv. farm 1.5 2.0 2.3 2.4 4.3 4.6 2.2 Cropping pattern Cashew nuts 0.22 0.21 0.44 0.39 0.92 0.58 0.40 Cereals - 0:13 0.18 0.46 0.11 0,67 0.24 Pulses-, rootorops - 0.26 0.22 0.42 0.58 0.63 0.32 Cereals/pulses 0.26 0.17 0.37 0.49 0.75 0.42 TOTAL 0.48 0.77 1.21 1.76 2.36 3.45 1.38 Labour input, hrs/farm 422 515 690 580 665- 1239 497 Household expenditures, shs 415 497 596 677 989 1,984 634 The Mtwara farmer'is 46 years old and his household consists of Z. people, including himself, forming a work force of 2.5 man-equivlanets. 15 percent of the farmers have off-farm employment and the remuneration amounts to shs 2,098/year for each of them. On average of all farms surveyed in Mtwara, the amount received would be shs 306. The farm size distribution, althouGh having both extremes of very small and large, is more equal than in most of the other survey areas. Farm sizes average 1.38 ha. 29 percent of this is under cashewnuts (of which 18 percent is inter-cropped with cassava and 14 percent with groundnuts). 17 percent 'f the cultivated area is used for cereals, mainly sorghum, rice, millet and maize and 23 percent for pulses, root and oil crops. Finally, 31 percent of the area is planted -to crop mixes of cereals and root crops or pulses, with mixes of rice, sorghui, maize,and cassava predominating. - 181 - TABLE A.5.2: Cropping Pattern in Mtwara Crop or crop mix Cultivation size, Percent on total ha/farm cultivation size I 1. Cashew nuts * 0.40 29.0 12, Cerals 0. 17.3 T Let 0.03 Sorghum 0.07 Maize 0.03 Rice 0.06 Maize millet 0.02 Rice/sorghum 0.03 3. Pulses, roots, oil crops 0.32 23.2 Cassava 0.17 Groundnuts/cassava 0.06 I Groundnuts 0.03 Cassava/beans 0.03 Sesame 0.02 Beans 0.01 4. Cereals, pulses, roots, oilorops 0-42 30.5 Rice/cassava 0.17 Sorgham/cassava 0.13 Maize/cassava 0.10 Maize/beans 0.01 Sorghum/beans 0.01 TOTAL 1.38 100.0. * 18 percent inter-cropped by sassava and to 14 percent by groundnuts. Rice seems to be the most important staple grain, and casaave the most important root crpp. For the average farm only 497 labour hours are used, with relatively little variation from farm to farm and between farm size groups (see Table A.5.1). Only in the largest farms of 2.50 ha and above is total labour input per farm more than doubled. These farms also have more than twice as much family members as the average which suggests that labour utilisation is a function of, apart from cultivation size and cropping pattern, labour avail ability. - 181.- Labour requirements for land preparation and planting, weed control and harvesting are summarized in Table A.5.3. Maize has, with 1,100 labourlhours/ha, the highest input and sesame with about 200 labour hours/ha, the lowest. All other crops and crop mixes lie in between this range. TABLE A.5.3: Labour Inputs in Major Crop Enterprisest Mtwara No. of Average Labour hrs/ha Crop or crop mix obserf- cultiv. Land prep bed Hves-o Total ations e ha planting control ting Millet 4 .39 211 48 69 328 Sorghum .78 116 56 181 243 Maize 6 .27 522 208 283 1,013 Rice 9 .30 28/s 156 170 610 Beans 3 .21 320 140 113 573 Cassava 17 .47 144 76 40 260 Maize/beans, groundnuts 7 .48 396 261 242 899 Cassava/beans 5 .32 434 242 214 890 Sesame 2 .43 87 67 54 208 Rice/cassava (maize) 19 *4,2 399 179 183 761 Cassava/sorghum 12 .52 259 203 116 578 Cassavalgroundnuts 6 .49 247 137 60 443 Cashewnuts 34 .54 53 168 206 427 For those crops or crop mixes for which complete and adequate observations are available, polynomial regression between cultivation size and labour inputs/ha were employed (Figure 1). For cashewnuts, rice/cassava and sorghum/cassava, labour intensity decreases with increasing cultivation size. For cashewnuts and rice/cassava, this drop is from 1,500 labour hours/ha on very small cultivation sizes, to about 500 hours at a cultivation size of 0.5 ha. The mix of sorghum/cassava has liwer levels of labour inputs. It should be remembered that labour is only one of several resource inputs, but is the only one which was recorded in Mtwara area. Household expenditures average 634 shs per household and are lower than in most survey areas. The bulk of it is spent on clothing. Sugar and stimulants also use major p*rtions (15.5 and 14.3 percent respectively). Household expenditures are summarized by item in Table A.5.2. � � � * •е I � �. . ' 'д .� у � н 1 + . .г.^ r й � В F ,.а й`.t.� ч - л+"< ` . ^'•" . , .л •. �- " ~ � b � , `.f't � . • . . . . . t.. ы"'.::, :: i. �7�..!' ... ' й'.да � J �. , ' } .�` .... '- < . . �` v. .. -Г _ , . _ е;. • у �. У .е7 . Т {р' - .. - �: � .. _ _ ,ь�У.-.ь.+мь.п..аwм •. а.п.-:-•т-'. �С с.,.._ьt з �-iы�+.+rь+iл.rь.�чл+:л`м�-.,;�-ыА+s.+псv-�ил..г�,<ге 4�aw'_,'и?%r .тv�ьr.....+л ...,sv.г �л.-сл.-.а.п+v �=..г+.4иСд. -и�-ь�.и.иnн�<"i' �•а-аи�мгV-.п- �^ '- ч .... и.•.._..... �.....б .. . .� � � ` � . , � • . !` ^' t. �' г . ^ 'SV - �ввь3йi.2..4а�д.К��_У[::л�л.W.ti.+�'1�.втtг,+_.йК...т+/�г.'ч�•.S•-`�L_. Л:•лТ+'.п�-:4w:�вiг.Кa:L�FF У • -` t ' � +�•��'�`3 s..�..R к�<r-.н-:x+r.<_:, � ь j[iурду' �.: usг+.p+Xr.1�a�Ya1r..�::^a�J..-, �аqS,уалr.r•гFлиг�Ч'к<rл�+t.aGC'lУ..ъ�;'l�.s+л.l�•'>п�...+^-уа л-�.Ср< �.sУаУ�-ь. �кL'*. ц,д i' � г � ,� , • w� �б.,.t�i и.,..w д г ' • �г;� [. г . � • . . г � � . ' рΡ гу � , � � , ц • 4 ц � . . �..G в+ У-х �' • � -т . {г Г _ � � � 5 У ; � F� Е'?it' !„ют !а':+., ,`' � с i 'r j � „ `rЬ,_•л t � . , � � � (-� � в " г r +"б�а�+ • . � � . . < , __ 'У А� ч�.�-i <� _> t� ` :. ; • �� � . . I �%. � �э '/s ' a-i и) Б ,.� .� ' . � ` - �' . . '' гг ; + � : 5 . . .. . , б _ . . ._ _ . ....: ..� �,..{< tr. �!. Ф�,� _.. " '• а+"3 ' . " .' ".' - �., _ _ ,' , ' ё . �� ч а'Я � � ( ' . � � � � У�•.�-~-..г-^г':' � � с' у � `; ` ���� '<й : З ? '`� ` + ,г i � ' -'�о ~-4 � С� .`.га � .._н9 � � - . � ( ,. i ^� Б е , � � , . . .. � , т: . : Г4! � jt ! . "'`�.'s . � S . . , - . _ . .�У'о .-. -. . - - i .. � _._ . . , _ .._ к -. .._ .' '. .- .' ' .. . . . . �� � .,::� �§ � ` �� � . . i . � �. . � . ; . ^�` .. _ . �� � � . .. � . . . � �,� � • ; , : : €�� k , � ; � - , • ' ' ' �� : z. рΡ G � , ' . _ : ' � , С. 2 . . . � .i • ' ' ' � �.) Yj ьд3 . � . - � ` �- ._ �. _. ._.._ .. �. ._. _,.. - ' ...,... ,_ Cr.:..L-г . . .-ь._. _'_у. "4' _.. -. _ .. _. . . и" ;� цi� � � �. j �^? �+ _�_ � � '� � •� :ч.! . � . _.. . .., .. .. '� "� .. �д � '. : � ... .. $ _ ! `,i •' _' ' �. "'_. ..._ 1 . '.. S .,... t ' . � ' . _ г- . � f. ��� ��• . Р1} � '} �! � f � � � � 1' wнF ' � i . - ; . . � . ' ... � ' . � � -� � • . :;' . . � .. fr� . У}д ; _ . . . "� . . _. --i ; , ; . . .�'--• - "+. : _ -��} . , .. . - � .._. _г_ .. . ._ ' ;. � �,..мь �'Э . �� . � �.�' . � „� � • : . ' у'а а C+i .� • . "� г . ... .. • _ '._.[4� � . � � .i• ... ..._ : .� _ ._�в3,� . . � г .. _ г � G' -й . . . r r � , . Г"`-7' • ' .. ._ - _. . ' . , . _ . •< . ' ' � �� ` ""зтобn-_`саь"°�°" � � � .. � � � �. ' � � � tn . . . .. . _ .. .� 1: j, .. . .. _ . _ : . • :ц� , � : , . Э , . , � . : . : -; .. . _._.. ; ' ., .: _... : . 4»7 ' р._.�л..•А,. _:t.rrae�^.�°_-_- -_ " ее+4глnчs'л�ir�sагвв . �ьм�глr..sпwта)� . • ' .ьу - - __�__ ' ' �� �'� " с ��> --+гGКы_$,*�s .� ..о:мF+лх'�.: •к'wo Ь• Й� � � . �`•�-� н �а �?� , .. . . �� . ; Аа� . • Е+..'r.:����и : �.� � �� �, � . . . � �- . . .. _ � - - `г. .� + _• v . % � • �" �i к •<,.., н \. r ��`б w�¢ 4 ki.�'`i �"` k � • �' •�ал � �1�•, '�а= • . • � г 6„ . .., ^ - �'го' . . • - . � v у'а` tr ,� А `�ч �г �г... � t:.. • &� т -• � ..f- . . . , � ; Т L< ' . � а � � t � G:аи=' ---л+s_..г.л-....�=л.•�.as-+..-:+w+r�т...м�=,r...и.wл..г��..�...._е,,,• . _ _.,,.. _ ! . - .�.мwин.->,r •.чмь.дяwс..:н+...иью:л.иигти.++а.�г...ео�.+меr:..гмwчм.+гт_.а+жlк�w•:.жь.л.ттг-rsv-и�r.,+жг�- - •�т..:.,-ц•a.wve.�,ra�..кv.a.:Gи.г.п.�_л,:r�:..гsт..л � . � � _ у.,,,. + + . .: � � г�i !"✓ V . .К�. _ � • . . . � . ' = ' _ ч м . а.+. • . и . � • � " ,� st- _и 'w Т 6�" ,л � ,. 2 у в2 , � �� �' . : i._ � 6 4.�г,Е_у в, �г- �,. � - � � г. ..� ,, .,° д. t " Q� г; '` . й , . t z jpi t_ Et'h- м �ъ_ у � ;�и'�,5„_L� �iс,� i.:+l� •"ro. а`у' _{ � `" 'Э "�,-'+ Е ��'�• S, � t�S; � . �' :` ).`�`" r�"!. `J'. _ъ <� :.б v.. v �•ь.�уΡr � . , S �` _ _м�Ч,рΡ_+ • �+J.. ! -я.. ь :i с .г_ , у � .... _ l' �� t ,н 25л + е" ' О t±^"а-`,у ....г � г`" ,ь� Е_е� •�, 4` 'т'3 L"•t r L „У д.� вТ "r е�. L�` и_✓ ,� �-� ,�� ..fi L. г-. ,.S с Si 66 <.л iц'� <<+у Ъ ы с л� �;а '-� С у . ц�� � �z * 4_ � . . . .. .. _ 1�. . М� � , У� .� __ .� . ._ _� � . - ---' - ' •,�' �' ' Q � � ' : _ �1 . � � , ( . , �� '�. �мt'-ii`4 . , � �. � .. . �ч.•�� r�` .- � . •� . ' � � ' i � � �� . ` . '-. . г . � t 4 � .��' rt . , .. ; м и .. . � (� ' • .', .+. �• - ' � , ` _ _ � . ' . . ` 4r � "- � � . г'� `''`: а:аЭ ��г ' .i ., .. . . : � �:�. г• ?�- - • • �. . � �х,л� �„+ . ' . � ' ' � _ : • . : � . } - Т - . . , _: к.л,•жн+w,м,..r•а�.....кцли........,..е..�v�л.--. ._.,в-+..д.•ч..~nл...,....гw.....с..«,.w.++д..++.v-..лwиr..гг.м<с.чм.+..ет�r.��л-т+.i.х�б-,.еrъзм..е.,.ы.ь.вw�.а-n...+.,иw«иа.+.ьч.+.�.....,чr�-... •У...,._.ту:-г,�+.:.... • .. _-г. _.••-•-_... �_-�м-+�✓r.:жисяsыт_+.�.:.�..-,�, � , а ' , . л ' у ` ` . � • ' ,• рΡ^�� . . .. . . . . , . . . • . .. . i ... . . . 4 � ..... _.. в � . • . . . .. {,рьq . . � '!д� _ 1 _ .. . . . -; . . . . , . . . . . - 184- TABLE A.5.4: Average Household Expenditures in Mtwara Expenditure item Average expenditures, shs Percentage on totall expenditure Clothing 257 40-5 Sugar 98 15-5 Stimulants 91 1 3 Tea 29 4'+'.6 Tobacco 62 9.7 Kerosine 56 8.8 Household goods Z' 3 6 7 Condiments 33 5:2 Transp:)rtation 21 3 3 Medicine 16 2:5 Alc. drinks 16 2 Others 3 0 4 TOTAL 63 100.0 A.6 Mwanza Area In Table A.6-11 basic farming characteristics from 37 Mwanza farms are summarized. The Mwanza farmer is 42 years old and has attended school f:)r only 3.3 Years. His household consists of 5.3 people -r a workf-)rce fcr agriculture of 3.4 man-equivalerits. The farm size averages 2.65 ha; and farm sizes are reasonably equally distributed. Crops grown are cereals, notably millet, maize and rice, which are grown on 1.15 ha or ,4 percent of the cultivated r-xea and are more d,-.minent in smaller farms. Cotton is grown on 0.66 ha or 25 percent nf the cultivated area. The area under cctt )n gradually increases with increased farm size. Finally, 0.78 ha are under ro t crops, -,)ulsos, oil crops or other crops and crc p mixes. - 185 - TABLE A.6.1 Characteristics of Mwanza Farms Farm size group, ha Details 00Average Under 1.00 2.00 - 3.00 - 4.00 - 5.00 & 1.00 1. 9 2.99 '3.99 4.99 above N o . of ob servations 7 --_ 7 1 6 5 1 37 Cultivation size, ha 0.52 1.54 2.41 3.35 4.45 8.95 2.65 % of farmers 10.8 18.9 37.9 16.2 13.5 2.7 1oo.o % of total area 2.1 11.0 34.5 20.6 22.7 9.1 100.0 Farmer and family Farmer's age 53.0 38.9 41.5 35.8 40.8 56.0 41.7 Farmer's e'ucation 0.7 2.3 4.5 3.0 "'2. 1 9.0 3.3 People/farm 3.7 5.8 5.1 5.7 5.6 8.0 5.3 Man-equivalents/farm 2.3 3.4 3.7 3.2 3.4 5.0 3.4 Cropping pattern, ha 1.07 1 Maize, millet, rice 0.33 1096 1.39 1.86 2.45 1.15 Roots, pulses, oil crops 0.11 0.05 0.62 0.79 0.78 - o.69 Cotton - 0.21 0.49 1,08 1.40 - 0.66 Others 0.07 0.21 -14 0.09 0.06 6.50 0.29 TOTAL 0.51 1.54 2.21 3.35 4.10 8.95 2.59 No. of cattle - - 1.0 - 6.2 - 1-4 No. of g3ats 2.3 2.3 0.8 0.3 2.6 - 1.4 Household expenditures, shs 1 1,149 1 1,444 1,586 1,140 1,345 838 1,523 1.4 head of cattle and equal number of goats are kept on average. 13 farmers keep goats, 34 animals per farm, and valued at 52 shs each. Information about household expenditures indicates that 1,523 shs are spent per fari, with no clear trend as t3 how these expenditures are related to other paraneters. oL - 186 - Expenditures per item are as follows: Item Shs % af total Clothing 750 -,r9.3 Sugar 269 17.7 Alc-holic drinks 175 11.5 Detergents 91 6.0 Kerosine 7, 4.8 Cooking oil 53 3.5 Medicine 44 2.9 Condiments 22 1.4 Others 45 2.9 TOTAL - 1,523 100.0 4.7 Mara Area From 60 farmers in Mara area, information about their basic farming situation is summarized in Table A.7.1. The Mara farmer is about 51 years of age and has attained 1.7 school years. His farm household consists of 6.5 people, making a work force for farming of 4.5 man- equivalents. The farm size structure in Mara area is most unequal. 60 percent of all farmers hold altogether only 26 percent of all cultivated land. A large number of small farms exist side by side with relatively large farms. Farm sizes average 1.85 ha, but this average means very little; farms are either much smaller -or much larger than this. About half -of the cultivated area is under cereals (millet, sorghum, or maize) and the other half is devoted to cassava, beans and sweetpotatces. Smaller farms tend to gr-w a larger proportion of the latter crops, while the larger farms grow care cereals. In addition to crop cultivation, livestock are kept in equal numbers on all farm size groups: 2.9 head of cattle, 3.9 sheep and 3.3 goats and 3 chickens. - 187 - TABLE A.7.1: Characteristics of Mera Farrus Details Faaa size r ha Under 1 0.50 - 1 1.00 - l 1.50 - i 2.00 - 12.50 & 0.50 0.99 1.49 ' 1.99 2. 49 laver No. of >bservations 13 3 6 5 8 15 60 Cultivation size, ha 0.34 0.69 1.13 1.62 2.17 4.34 1.85 % f farmers 21.7 21.7 10.0 8.3 13.3 25.0 100.0 8.2 6.2 7.3 15.6 58.6 100.0 Farmer and family Farmer's age 49.8 48.6 e6.0 29.2 149.4 54.2 50.6 Farmer's education 1.5 1.2 3.0 1.6 2.1 i 1.6 1.7 People/farm 5.2 6.6 5.2 6.4 5.5 8.9 6.5 Man-equivalents 3.4 5.0 4.0 3.9 3.3 6.1 .5 Cropping pattern, ha Millet, sorghun, maiz 0.15 0.24 0.28 0.52 1.14 2.53 0.95 Beans, cassava, sweetpotatcLs 0.19 0.44 0.75 1.10 0.91 1.39 0.79 TCTAL 0.34 0.68 0.13 1.62 2.05 3.92 1.74 No. of cattle 1.4 3.2 1.7 6.0 1.8 3.8 2.9 No. of sheep 1,1 0.8 19.8 2.6 1.4 4.4 1 3.9 No. of goats 1.9 2.6 4.0 2.6 7.4 2 3.3 No. of poultry 0.5 4.4 2.2 4.6 2.9 3.8 30 - 188 - APPENDTX B SUPPLEMENTARY TABLES SHOWING GROP LABOUR REQUIREMENTS BY REGION Supplement to Table 2.1 Labour Distribution by Major CrQps pur Farm - Kilimanjaro - CDffe-Banana hrs/ha Maize-Beans hrs/ha Farm hrs Average Best Averape Best Distribution by seaso)n Jan 1 31 83 170 126 Feb 29 31 36 1',5 122 Mar 89 125 76 301 181 Apr 75 216 33 65 153 May 86 1,5 13 - 153 Jun 59 215 132 508 18 jul 76 82 26 21 192 Aug 206 313 - - 278 Sept 1438 - - 230 Out - 23 .;21 2 29 3 Nov 13 31 39 160 148 Dec 14 31 121 i 310 183 TOTAL 825 2,078 559 1,709 2,093 Distribution by type of labour Fanily 763 1,902 545 1,688 2,017 Seas. hired labour 62 6176 14 21 7 TOTAL 825 2,078 559 1,709 2,093 Distributi'n by operation Land preparation, planting - 279 84.7 Care 255 351 111 338 Harvesting, others 570 1,727 169 524 T4eTAL 825 2,P78 559 1,709 - 189 - Supplement to Table 2.6: Labour Distribution by Majzr Cr'xns per Farm - Lushotn - Distribution (hrs/ha)Far by season Maize-beans Maize Veget- COffee- Coffee jables banianas Average Best Ave rage Best Jan 48 177 103 204 37 103 - 166 Feb 51 226 3 62 3,; 39 j 57 147 Mar 119 134 27 98 28 99 114 150 Apr 22 78 22 141 14 19 1 109 130 MAy 23 12 82 72 89 5 - 1'1 Jun 20 19 6 157 25 26 21 122 Jul 30 74 55 252 97 147 123 173 Aug 81 431 7 172 31 193 133 178 Sep 14 18 2 16 89 51 77 118 Oct 24 97 1 - 51 63 73 123 Nov 28 131 11 94 - 48 - 121 Dec 226i 4 - 93 47 137 I 1 7 TOTAL 424 1,614 i 323 1268 587 837 7 1706 Distribution by type -- labour Family 424 1,614 325 1,26 587 837 709 ,706 Seas. hired labour - - - - - TOTAL 424 1,614 325 1,268 587 837 709 1,706 Distribution by operatijn Land 173 612 146 567 242 - care 143 567 12573 193 457 3 4- ithers 103 - 435 52 222 152 380 366 TOTAL 424 1,614 323 1,268 587 837 709 Supplenent to Table 2.12: Lanbur Distributiz n by Maj,r Cr )s an,3 y Far-I - Mbeya, hrs/ha - Maire Wheat Pyre- COffee Cjffee/ Beans Peas Maize/ gillet Farm Average Best thrum banana, b ean s (30) (6) (16) (13) (9) .(3) (29) Distributi-n by scas-n Jan - - 12 6 182 103 - - - - 42 Feb 58 119 17 26 15 67 38 17 56 18 75 Mar 23 78 11 32 64 7 16 29 22 1 :3 Apr 6 - 1 21 141 17 5, <3 - id 44 May 8 - 6 31 - - 38 26 - 5 18 Jun 135 213 51 <0 96 2 25 9 22 45 134 Jul 4 - 9 89 20 57 66 17 24 - 39 Aug 11 25 16 58 27 67 - - 10 - 36 Sep 21 65 28 5 13 20 - 26 - 14 27 Oct 12 8 15 11 - - - 34 32 - 18 Nv 67 57 7 29 - - 13 - 46 9 65 18 i,8 32 32 45 10 22 43 29 41 42 TOTL 363 653 206 380 603 350 271 245 240 155 582 Distributien by typei of lab-ur Family 332 508 203 240 596 550 271 236 240 150 520 Seas. hired labour 31 45 3 140 7 - - 9 - 5 62 TOTAL 36, 653 206 380 603 350 271_ 240 155 582 Distribution by -perati.:n Land preparatic(n, planting 135 245 110 178 - - 105 125 112 91 Care 82 187 29 65 ;33 205 75 69 83 14 Harvesting, thers 1/6 221 66 137 170 145 91 52 45 50 TOTAL 363 653 206 380 603 350 271 245 240 155 -27 -4 40 rTO-Tli Slonent to Table 2.16: Lab.ur Distribu-ti by Ma;j,r Crops and perFrV - C-astal area, hrs/ha - Cashewnut/ Cashoimut Co:c,nut Rice Rice- Maize Cassava* Farm hrs C cnut _aize Average Best Average Best Average Best Average Best Jan 25 20 3 14 20 27 142 16. 28 12 56 119 Feb 19 19 6 23 20 27 99 138 162 54 42 104 Mar 22 41 3 14 20 27 94 178 133 284 42 107 Apr 18 21 3 14 20 27 2.3 393 Al la 62 149 May 29 52 3 14 30 42 138 280 125 96 60 141 Jun 57 18 12 l, 38 27 96 41 54 -149 72 178 Jul 38 77 71 198 20 27 77 264 114 170 119 133 Aug '5 51 51 82 20 27 69 321 11 12 42 154 Sep 76 122 225 583 70 108 42 52 11 12 42 315 Oct 34 3 5 24 118 89 36 4l 75 128 56 115 Nov 35 22 9 22 24 34 64 251 4l 254 58 121 Doc 29 39 3 14 20 27 18 49 78 281 ^5 93 ___ 427 485 393 1,007 422 4 1,117 2,172 374 1,jl7 697 - 1,782 Fa.ily 338 392 313 1,007 304 :47 1,025 2,054 814 1,354 684 1,199 Seas. hired labour 89 93 80 - 114 42 92 118 60 263 13 282 TOTAL - 27 485 33 1,007 422 <89 1,117 2,172 871 1,617 697 1,782 Land prop. 103 137 - - - - 382 352 380 752 3i3 Weeding 118 118 151 399 393 402 198 329 146 - 7; 22/ HarvestinG 206 230 2<2 608 29 87 537 991 238 391 160 TOTAL .,27 -85 393 1,007 :i22 489 1,117 2,172 874 1,618 697 ipplement to Table 2.22: Labour Distribution by Major Grops and per Farm - Morogoro , hrs/ha hrs/ha- hrs Maize Rice Sorghum Maize/Sorghum C-tton Farm Avge Best Avge Best Mecha- Auge Best Avge- Best Avge Best nized ribution by season 215 62 456 3,057 135 191 134 110 63 74 16 224 266 37 206 477 190 293 515 241 396 .423 389 230 242 418 278 624 35 194 493 273 364 336 406 216 333 17 350 509 416 299 462 554 1,030 192 262 231 330 241 524 886 521 515 554 79 161 131 196 389 164 143 398 1,174 194 94 138 89 - 93 184 156 116 188 L49 - 70 2,:,0 218 159 311 - 112 116 149 18 81 - 74 - - - 618 -818 106 73 - 39 - - 19 - - - 659 308 225 57 -143 A8 - 14 154 562 14 - 28 - 67 156 169 200 554 10 133 289 203 321 276 301 157 149 155 145 65 83 258 - 296 47 0 79 15,' L 2,107 1,722 2,811 7,627 1,668 261 3,365 2,018 116 2,870 2,959 1 2,267 ribution by type of labour ly 1,903 1,628 2,367 6,671 1;6o6 2,425 3,365 1,947 3,07,-" 2,833 2,920 2,067 .hire-d labour 214 9 444 756 62 36 - 71 42 39 200 L 2,107 1,722 2,811 7,427 1,668 2,461 3,365 2,018 3,116 2,870 2,959 267 ributicn by operation . proparation, ant in g 736 720 949 3,757 205 7 4'9 1,003 599 99 542V6, in, 574 363 487 1,378 297 579 989 521 745 1,110 1,340 -est.. 797 639 1,375 2,292 1,166 1,133 1,373 0 1,413 1.216 1,001 1 2,107 1,722 2,811 7,427 1,668 2,461 3,365 12,018 3,116 2,87C 2,95c , Supplement to Table 2.32: Labour Distribution by Major Crops per Farm - Shinyanga - Hrs/ha Hrs Maize Maize/ Cotton Rice Sorfhum Maize/ Sweet Cassava Ground- Sorghum Groun&- Potat- nuts Farm nuts oe Distribution by season Jan 34 14 75 - 33 33 37 37 71 136 Feb 105 95 103 53 113 116 63 145 49 282 Maar 35 26 122 27 15 22 40 79 71 162 Apr 2 2,, 13 11 - 28 26 37 - 7 May 1/9 1/;5 145 103 29 29 103 63 84 344 Jun ', 29 156 87 1.2 94 62 35 3o 211 Jul - - - - - - 7 - - 44 Aug - - - - - - - 31 - 46 Sep - - - - - - - - - 43 Oct - 9 36 - 15 - 51 135 3-- Nov 56 11. 48 151 22 72 135 44 15 245 D0e 141 95 153 307 140 1659 262 181 106 ,15) TOTAL 562 551 357 749 508 564 790 790 .,75 2,131 Distribution by type of labour Family 194 551 545 569 501 503 690 721 475 11813 Seas. hired labour 68 - 312 180 7 61 100 69 - 283 TOTAL 562 51 87 57 7...9 508 564 9 790 97 2,131 Distribution by operation Land preparation 72 130 84 293 32 123 306 224 15 Planting 125 76 153 165 118 114 165 135 106 Care 175 143 314 94 160 172 78 150 190 Harvesting 147 136 222 192 115 125 235 225 164 Others 3 79 2 83 26 25 26 - TOTAL 56 2 357 7!1 508 564 790 790 475 Supplement to Table 2.33 Labour Distribution by Major Crops (Best Productivity and Different Mechanization Levels) - Shinyanga - Hr /ha Maize Maize'r Cotton Rice SorChuma Oxs Tractor Ox Tractor Best Trautorbest ploughed ploughol Best Distribution by season Jan 12 17 54 20 91 69 - 157 32 63 160 131 87 121 Mar 66 26 43 9 - 135 20 Apr 13 3 - 52 - 23 - May 118 102 193 132 1`3 203 161 Jun 150 47 12 39 279 56 to Jul - - - Aug----- Sep - 7 6- Oct - - 20 7 63 Nov 8 5 23 169 - 101 0 Dec 165 78 228 53 159 175 472 TOTAL 639 360 621 709 893 922 851,. Distribution by type of labour Family 677 343 609 709 701 879 71' Seas. hired labour 12 12 12 . - 197 :3 136 TOTAL 689 360 621 709 898 9224 Distribution by oerg.tion Land preparation 153 110 21 31 167 376 52o Weed control 237 171 130 150 345 233 136 Harvesting and others 299 79 200 241 386 313 198 TOTAL 689 360 621 709 398 922 854 Supplement to Table 2./0: Labo)ur Distribution by Major Crops and by Farm Tabra hrs ha hrs Maize, Groun- Maize Maize, Cassava Cass, Sweet- Tobacc.9 Farm nuts, Cassava ava 2tr-o1es Avge Best Fort Avge Best Fort Avge Best Fort Avge Best Avw- Bes AG _ 1de Be -- Bes Distribution by soason Jan - - - - - - 58 113 115 168 71 172 "1 80 95 Feb 109 15': 102 141 161 1.1 156 208 270 156 165 176 152 180 207 Mar - - - - - - 44 26 - 23 32 - 22 - 71 Apr 213 431 280 86 55 36 95 285 392 - - - - - 224 May 112 133 118 119 180 112 142 157 - 119 - - 567 702 229 Jun 34 33 38 49 6 43 65 - - 22 32 100 106 138 112 Jul 5 - 8 - - - 23 38 - 12 - - - - 71 Aug 12 0 12 4 - 5 30 86 212 2 - - 52 - 79 Sop 10 9 6 24 27 20 24 70 136 26 19 --154 119 oct 3. -' 27 39 58 35 64 76 - 57 34 260 240 333 133 Nov 15 5 18 8 - 8 53 27 - 29 - - 198 320 108 Dec 149 213 156 173 207 164 93 100 125 49 113 6.1 210 278 237 TOTAL 699 162 765 644 752 614 8,7 1186 12.o 662 518 772 172 2200 Family labour 697 L162 762 <89 752 614 8,,'7 1136 12,0 518 772 1151 903 1589 Seas. hired labour 2 - 3 155 - - - - - - - 591 1297 70 TOTAL 699 162 765 6.4 752 61. 847 1186 1240 662 513 77 2 . 220 1 Distribution by operation Land preparation, planting 211 288 241 2.'9 292 241 252 310 394 246 346 96 538 721 ffeeding 1 4 140 121 141 161 1.42 180 171 167 126 139 176 303 362 Harvesting and >thers 344 63/ 403 25.: 299 231 415 702 679 290 33 100 852 1117 TOTAL 699 £162 {65 6 7 614 847 1186 1240662 518 772 172 2200 AvGe - Average Fort = fertilized Supplerent to Talle 2,4.9 Labour Distribution by Maj-r Crops - Iringa - Hours/ha Hrs Distributionby season Maize/beans/ vegetables Maize Millet Maize/Irundluts Fr (cr9ps) Average Best Average Best Average, Best Average Besnt Jan 63 118 22 6 - - 39 24P 100 Feb 42 56 36 52 4 104 75 155 99 Mar 18 41 8 - - - 8 - 5i Apr 5 13 - - 63 167 - - 36 May 25 25 - - 19 21 13 -53 jun 38 65 15 23 54 118 117 Jul 37 80 21 23 - - 11 24. 69 Aug 32 1 10 12 3 - 20 24. 55 Sep 3 9 - - - - - - 24 Oct 12 - - - - - - 22 Nov -3 - - - 23 Dec 72 91 39 67 51 8 110 275 119 TOTAL 3 3 60^ 150 177 237 91 391 7: 735 Distribution by type of labour Family 3.'1 60/ 131 1..:3 237 9.: 39 78 735 Seas..nal labour 2 -19 29 - - - TOTAL 313 604 150 177 237 9 39 7735 Distribution by onoratiun Land prepiration/-lantinU 153 263 63 67 122 201 1-3 23.. Care 68 96 3 2 - 78 37 220 Harvesting, othe,rs 121 2 72 115 16. 29 TOTAL 3 2 60, 10 177 237 9 39p 7-8 . ï . Supplement to Table 2.57: Labour Distribution by Mal r Crops - Ruvuma, hrs/ha -7 Hrs/ha hrs Maize/Beans Maize Maize/ Tobacce Coffee Cass- Snrg- fillot Rice Beans Farn Cassava Millet ava hum Avk Best Fert Avg Best Avg Best Avg Best Avp Avr Avg AvgI Avg Avg Avg ,Dijtributicn by season Jan 141 125 347 80 15,; 76 220 122 57 36 12 31 23 - - 29Q Feb 65 133 - 20 - 177 254 89 102 56 72 14 299 383 40 329 Mar 9 - - - - 145 56 346 380 64 5 116 - 1436 56 340 Apr - - - - - - - 17 - 6 72 - - - 7 118 May 37 41 204 13 7 9 - 521 635 28 7 - - 105 19 220 Jun 108 225 7 135 165 163 330 297 224 371 30 80 252 166 24 444 Jul 28 - - 15 - - - 21 - 352 51 86 - 15 23 212 AuE 33 47 - 6 25 8 33 11 - 673 28 147 - - - 228 Sep 31 54 - 23 35 50 20 41 27 13 103 30 - 68 - 225 oct 33 24 - - - - - - 22 62 14 177 - - 171 Nov 97 105 127 42 101 131 117 101 112 18 67 2 8 165 38 341 Dec 64 71 124 88 3, 40 54 55 47 5 40 61 129 8 96 291 TOTAL 658 825 809 423 521 797 1104 1622 1614 1686 550 501 888 2527 303 3218 Distribution by type of labour Family 6 ;6 769 752 368 521 797 1104 1607 1614 1662 550 308 888 2527 303 3122 Seas. hire, labour 12 56 -"7 5 - - - 15 - 24 - 273 - - 96 OTAL 658 825 309 423 521 797 1104 1622 1614 1686 550 581 888 2527 303 3216. Distribution by operation Land prep, planting 271 321 254 174 206 240 267 324 215 - 25 290 337 444 197 Care 223 23 395 101 150 285 514 486 539 155 123 133 299 1023 40 Harvosting, others 191 270 160 118 165 172 323 812 86c 1531 173 158 252 1061 66 l TOTAL 635 825 309 423 521 797 1104 1622 1614 1686 550 581 888 2527 303_ Avg = Average FErt = FertilizecI 198 Supplement to Table 2.64: Lab-ur Distribtion by Major Crops - Singida - Hrs/ha Hrs Details Millet-Maize-Sorghum Maize-Millet-Groundnuts Farm Average Best Average Best Average No. of observations 56 17 23 6 Distribution by season 10 + Jan 20 40 7 83 244 Feb 32 60 42 39 232 Mar 7 9 29 75 178 Apr 1 2 3 - 140 May 64 117 49 38 345 Jun 11 12 31 116 168 Jul 11 7 15 22 176 Aug -- - - 109 Sep 33 69 13 - 225 Oct 19 29 19 42 189 Nov. 2 3 5 1 124 Dec 46 73 28 18 279 TOTAL 246 421 279 434 2,460 Distribution by operation Land preparation, planting 69 97 85 O5 Weeding 91 181 94 171 Harvesting 86 143 100 178 TOTAL 246 421 279