THESyc This is to certify that the thesis entitled ENTERPRISE CHOICE, ENTERPRISE COMBINATIONS, AND INCOME DISTRIBUTION AMONG FARMERS IN SIERRA LEONE presented by Steven C. Franzel has been accepted towards fulfillment of the requirements for M.S. Agricultural Economics (OW/“ti“ Major professor degree in 03639 OVERDUE FINES ARE 25¢ PER DAY PER ITEM Return to book drop to remove this checkout from your record. w .7» P ENTERPRISE CHOICE, ENTERPRISE COMBINATIONS, AND INCOME DISTRIBUTION AMONG FARMERS IN SIERRA LEONE By Steven C. Franzel A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Economics 1979 ENTERPRISE CHOICE, ENTERPRISE COMBINATIONS, AND INCOME DISTRIBUTION AMONG FARMERS IN SIERRA LEONE By Steven C. Franzel This thesis provides information on how poor rural house- holds select and combine their activities, so that rural development planners can direct their programs towards those in greatest need. The analysis is disaggregated by income groups to explore conditions of poor households and differences in enterprise choice and factor use among income strata. There is considerable inter-strata variation in enterprise mix. However, this variation does not contribute significantly to inter-strata income differentials. Rather, it is likely that dif- ferences in resource productivity in the principle enterprises are important in explaining disparities in incomes. The most important production systems are modeled by iden- tifying principle enterprise combinations. The analysis establishes the importance of peak-season labor bottlenecks and capital scarcity as constraints to increasing the incomes of the poor. The policy recommendations arising from the analysis empha- size increasing labor productivity in upland rice production and reducing peak-season labor bottlenecks. Furthermore, farm models must be disaggregated by income strata so that programs can be specifically targeted to reach low-income households. ACKNOWLEDGMENTS I wish to express my thanks to the individuals who have assisted me with this effort. Peter Matlon, my thesis supervisor, provided me with invaluable guidance throughout the course of this study. Derek Byerlee's aid in acquainting me with the data base, planning the scope of the thesis, and in reviewing my thesis draft is gratefully acknowledged. Carl Eicher, Carl Leidholm, and Warren Vincent offered helpful comments and support. Thanks go to Pat Eisele and Janet Munn for their typing and administrative assistance. I am also grateful to the staff of Agricultural Economics Computer Services for their aid in data analy- sis. Among my fellow graduate students, I wish to thank Steven Haggblade, Karen Klonsky, and Stephen Davies for their insightful comments. Finally, I am grateful for the support provided by the Department of Agricultural Economics. ii LIST OF LIST OF Chapter I. II. III. IV. TABLE OF CONTENTS TABLES FIGURES INTRODUCTION l.l Problem Statement l.2 Objectives l.3 Source of Data SIERRA LEONE: GENERAL CHARACTERISTICS AND THE RURAL ECONOMY . . . . . . . . 2.1 General Characteristics . 2.2 Agriculture and the Rural Economy. 2.3 Agricultural Policy in Sierra Leone 2.4 Income Distribution in Rural Sierra Leone ENTERPRISE EMPHASIS IN RURAL SIERRA LEONE . 3.l Enterprise Emphasis--A Descriptive Analysis 3.2 Enterprise Returns . 3.3 Differences in Enterprise Emphasis Among Income Groups . 3.4 Comparative Expected Net Margin Analysis: Returns to Labor and Land . . . 3.5 Enterprise Combinations FACTOR USE AND ENTERPRISE EMPHASIS Seasonal Labor Use Among Regions Enterprise Variable Costs and Capital Costs Comparative Expected Laobr-Land and Capital- Labor Ratios . . . . . . . . . h-b-bb th-J iii Seasonal Labor Use for Individual Enterprises : Page vii Chapter Page V. FACTOR USE AND ENTERPRISE COMBINATIONS . . . . . 9l 5.l Seasonal Labor Use and Enterprise Combina- tions . . . . . . . . . . . . . . 92 5.2 Enterprise Combinations and the Use of Hired Labor . . . . . . . . . . . . . . l09 5.3 Peak Periods for Different Income Groups . . l24 VI. SUMMARY OF FINDINGS AND IMPLICATIONS FOR POLICY . . 127 6.l Summary of Findings . . . . . . . . . l27 6.2 Policy Implications . . . . . . . . . l34 iv TABLE 1.1 2.1 2.2 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 LIST OF TABLES Ecological and Demographic Characteristics of Rural Resource Regions in Sierra Leone . Rural Incomes Per Consumer Equivalent (ICE) in Regions of Sierra Leone . Rural Incomes Per Consumer Equivalent (ICE) by Income Group in Sierra Leone . . . The Importance of Different Enterprises in Land Use, Labor Absorption, and Income Generation in Rural Sierra Leone Farm and Nonfarm Enterprises Classified According to Net Returns Per Manhour . . Net Returns Per Acre for Major Crops in Sierra Leone . . Age/Sex Weights Used in Calculating Consumer Equiva- lents . . . . . . . . . Enterprise Emphasis Among Income Groups in Rural Sierra Leone . . . . . Enterprise Emphasis Among the Highest and Lowest- decile Income Groups of Rural Sierra Leone Expected Household Net Margins to Labor and Land Among Regions of Rural Sierra Leone . . . Expected Household Net Margins to Labor and Land by Region and Income Group . . . . Major Enterprise Combinations of Rural Households in Sierra Leone . . Contributions of Component Enterprises in Enterprise Combinations to Total Household Labor and Income Enterprise Combinations and Income Levels of Rural Households in Regions of Sierra Leone . V Page 12 21 22 25 28 3O 31 33 35 38 4O 45 47 49 Table 4.1 Peak Periods and Slack Periods for Selected Enter- prises in Rural Sierra Leone Conventional Returns to Labor and Returns to Peak Season Labor for Major Enterprises in Rural Sierra Leone . . . . . Months of Peak Season Labor Conflicts Between Rice and Selected Enterprises in Rural Sierra Leone Annual Variable Costs and Capital Costs Per Acre for Major Farm Enterprises in Rural Sierra Leone . Annual Variable Costs and Capital Costs Per Manhour for Major Enterprises in Rural Sierra Leone Capital-Labor Ratios for Major Enterprises in Rural Sierra Leone . . . . . Cash Variable Costs Per Manhour for Major Enter- prises in Rural Sierra Leone Labor-Land Ratios for Major Farm Enterprises in Sierra Leone . . . . Expected Labor- Land and Capital- -Labor Ratios by Region in Rural Sierra Leone . Expected Labor- Land and Capital -Labor Ratios by Income Group . Slack and Peak Periods of Labor Use for Major Enter- prise Combinations in Rural Sierra Leone Hired Labor Use in Major Enterprise Combinations in Rural Sierra Leone . . . Hired Labor Use by Income Group for Selected Enter- prise Combinations in Rural Sierra Leone Variations in Peak Labor Periods by Income Group and Enterprise Combination in Rural Sierra Leone vi Page 63 67 7O 73 75 77 79 84 85 86 98 116 122 125 Figure 1 .1 LIST OF FIGURES Sierra Leone Rural Resource Regions and Selected Enumeration Areas . . Physical Regions of Sierra Leone . Monthly Distribution of Labor Used Per Household by Region in Rural Sierra Leone, l974-l975 Monthly Distribution of Labor Used for the Produc- tion of an Acre of Upland Rice in Sierra Leone, May l974-April 1975 . Monthly Distribution of Labor Used for the Produc- tion of an Acre of Inland Valley Swamp Rice and an Acre of Mangrove Swamp Rice in Sierra Leone, May l974-April l975 . . . . . Monthly Distribution of Labor Used for the Produc- tion of an Acre of Riverain Rice and an Acre of Boliland Rice in Sierra Leone, May l975-April l975 . Monthly Distribution of Labor Used for the Produc- tion of an Acre of Fundi and an Acre of Groundnuts in Sierra Leone, May l974-April l975 . . Monthly Distribution of Labor Used for the Produc- tion of an Acre of Cassava and an Acre of Onions, Peppers, and Tomatoes (OPT) in Sierra Leone, May l974-April 1975 Monthly Distribution of Labor Used for the Produc— tion and Processing of Wild Oil Palm Products in Sierra Leone, May 1974-April 1975 Monthly Distribution of Labor Used for the Produc- tion of an Acre of Coffee and an Acre of Cocoa in Eastern Sierra Leone, May l974-April l975 . Monthly Distribution of Labor Used Per Household in Small-Scale Fishing and Processing Production in Sierra Leone, May l974—April l975 vii Page 11 15 52 54 55 56 57 58 59 6O 61 Figure 5.1 5.2 5.3 Monthly Distribution of Labor Used for Small-Scale Industrial Firms in Rural Sierra Leone, May l974- April l975 . . Monthly Distribution of Labor Among Households of Selected Enterprise Combinations in Rural Sierra Leone . . . . . Monthly Distribution of Family Labor and Hired Labor Use Among Households for Major Enterprise Combina- tions of Rural Sierra Leone . Monthly Distribution of Hired and Family Labor Used by Households of Different Income Groups for Selected Enterprise Combinations . viii Page 62 93 111 119 CHAPTER I INTRODUCTION l.l Problem Statement Rural development planners in the Third World are becoming increasingly aware that information about the small farm system and its allocation of resources is a prerequisite for formulating success- ful rural development programs. The widespread failure of large— scale, capital-intensive agricultural projects and the increasing concern for a more egalitarian distribution of benefits have led to increasing emphasis on small-farmer oriented programs. The mobiliza- tion and increased productivity of underemployed resources, especially labor, in the small-scale rural sector, is seen as an important avenue to increasing output, employment, and income. Interventions in the small-farm sector have, unfortunately, not achieved the level of success that many planners once envisaged. The limitations of the widely-heralded Green Revolution high-yielding varieties are well documented (Wharton, l969). And even where Afri- can rural development programs have resulted in large increases in output and income, the low-income groups have rarely benefitted to the degree expected by program planners (Lele, l975). A large part of the problem is the poor understanding of how the small farm system works and how its resources are allocated. Often, planners simply ignore farmers' conditions and try to impose new "innovative" systems or subsystems on the small farmer which have been proved to be effective on research stations. The result is usually failure. On the other hand, the approach used in this study, the farming systems approach, views the farm as a system and examines the effects on a given changein acomponent (a new crop, for example) on the overall system. The immediate goal is to modify the existing farming system to incorporate improvements, not to replace the system with a new one. The focus is on overcoming the critical constraints which the farmer faces so he can increase his output and income (Norman, 1976). The large amount of information necessary and the location specificity of the farming systems approach are evident. The researcher must thoroughly understand the physical and biological conditions which the farmer faces, the resources at his disposal, his goals, his attitudes, and the decision-making framework which he uses to allocate resources. Knowledge about community and institu- tional factors and the pattern of resource use throughout the year are also crucial (Norman, 1976). Because the distribution effects of rural development projects now receive high priority, it is no longer adequate to use the aver- age farm in an area as a model for that area. What are needed, as was noted as early as l925 by the Russian economist Chayonov, are farm models for low-income households so that programs can be spe- cifically designed and targetted to meet their needs (Chayonov, 1925). A detailed study of enterprise choice is a vital component of the farming system approach. It is important that planners under- stand the nature of enterprise emphasis among different income groups. This information can enhance the understanding of why the poor are poor, and thus help in designing programs to increase their incomes. One important hypothesis to explain why the rural poor are poor is the following: low-income households have low incomes because they undertake low-returns enterprises relative to the rest of the rural population. The extent to which this hypothesis is or is not true may importantly influence the types of programs which are appropriate. If the hypothesis is true, programs targeted to help the rural poor should emphasize the introduction of new enter- prises to low-income households, especially enterprises currently being undertaken by middle and high-income households in the same area. Alternatively, incomes may be increased by upgrading the technology of low-income households in the enterprises they are pursuing. This latter approach may present greater difficulties, however,because of the uncertainty about the adaptibility of such changes. Epstein, for example, found farmers in South India to be more amenable to the introduction of new crops than to adopting recommended changes on traditionally-grown crops (Epstein, l962). If the hypothesis is not true, that is, if the choice of enterprise is not an important cause of poverty, other factors need to be investigated. One alternative hypothesis might be that low-income households have low incomes because of an inability to follow correct management practices due to constraints acting upon them. Another might be that differential access to resources (e.g., land) is largely responsible for income differences. Unfortunately, the data available to confirm or reject the hypothesis that enterprise choice is a significant determinant of income status is sparse and somewhat contradictory. Studies by Betu and Upton in the middle l960's in the savannah zone of Nigeria showed that the enterprise mix had little if any effect on gross margins per man-day (Petu and Upton, l964). Matlon, who studied income distribution in three villages in Northern Nigeria, found similar crop mixes among different income strata and that differ- ences in crop mix accounted for little variation in returns to labor or land. However, high-income households were able to give greater emphasis to several high-return "specialty" crops requiring high levels of purchased inputs. In addition, the percentage of income earned in the nonagricultural sector rose from ll.5 percent for the lowest-decile households to 35.4 percent for the households of the highest decile. For high—income households, the greater emphasis on nonfarm enterprises resulted in higher returns to labor (Matlon, l977). This study hypothesizes that enterprise choice plays a central role in determining income levels. If this hypothesis is true, the reasons why poor farmers have chosen low-returns enterprises must be explored. Some major reasons why poor farmers might reject particular high-return enterprises include: l. Capital constraints. Only wealthy farmers are able to meet the capital requirements necessary to pursue the enterprise. Even when cash requirements are small, they may come at a time of the year when cash is particularly scarce. 2. Risk. High-return enterprises may have high variability in returns from year to year, presenting the low-income farmer with a degree of risk which he finds unacceptable. The primary causes of variability in returns are high yield variability and high product-price variability. 3. Labor bottlenecks. Poor rural households may lack the labor necessary to meet the labor requirements of peak periods, especially when the peak period for the high-returns enterprise coin- cides with a peak period for the production of a food staple. Because of the cash constraint, or the unavailability of labor during the peak period, they are unable to hire laborers to meet the need for additional labor. 4. Health. Nutritional problems or disease may prevent poor households from undertaking high-return, labor—intensive enter- prises, especially when crucial labor inputs are required just before the harvest period. This constraint is related to "3" above. 5. Land scarcity. If farmers give first priority to meet- ing their subsistence food needs, it is possible that only the residual land (if any) will be left for cultivating high-return cash crops. 6. Exogenous factors which include: a. ecological characteristics--rainfall, soil fertility, etc. The poor may be concentrated in areas lacking the eco- logical conditions necessary for growing high-return crops. b. locational characteristics--area-specificity of some enterprises, proximity to markets, etc. Many high-returns enterprises, such as fishing and mangrove swamp rice, are area-specific. Where the poorest farmers live far from marketing outlets, the cultivation of cash crops may be unprofitable due to transport problems. c. demographic characteristics--age of farmer, family size, etc. Large families may be engaged in more labor— intensive enterprises. Older farmers may shun enterprises with heavy labor requirements. d. education—-awareness of new enterprises, contact with extension workers, etc. Farmers who can read and write and have contact with extension workers may have greater access to information concerning new high—return enterprises being introduced in the area. The relative importance of each of these factors has an important bearing on the design of programs for increasing the incomes of poor rural households. For example, if cash constraints are an important problem, the establishment of a credit program may be specifically called for. A credit program might also aid in the solution of labor bottlenecks. If hired laborers are simply unavailable at peak periods, however, changing the production period for one of the enterprises (by introducing an early-maturing variety, for example) or changing the enterprise combination may be necessary. Where increased risk impedes poor farmers from adopting high-return enterprises, programs which lessen risk are called for, e.g., intro- ducing enterprises with little variability in yields or prices. Alternatively, a program may help the farmer accommodate to risk by giving him access to credit. Although the above discussion has been based on enterprise choice as a determinant of income, the examination of enterprise combinations is also an important component of the farming systems approach. As mentioned previously, the researcher must understand how the introduction of a new enterprise (or modification of an existing enterprise) will affect the overall farm system. This can most easily be done by studying the two or three major enter- prises in combination, which serve as a representation of the system. An examination of the levels and timing of factor use in enterprise combinations provides guidance for releasing constraints to increas- ing rural incomes. For example, the peak season labor requirement is a major constraint to increasing rural incomes in Sierra Leone (Byerlee et al., l977). An analysis of seasonal labor use for major enterprises in combination provides guidance for adding new enter- prises or technologies to a combination, or substituting new enter- prises for existing ones. In summary, information about enterprise choice, enterprise combinations, and the levels and timing of factor use for different enterprises and combinations is vital for planning rural development programs. This information must be disaggregated by income group so that it may be used to design programs to improve income distribu- tion and aid those in greatest need. An analysis of enterprise choice among the rural poor and factors affecting enterprise choice can help planners target their programs towards low-income house- holds. l.2 Objectives The objectives of this study are the following: a. Describe differences in enterprise choice among rural households of different areas and income groups of Sierra Leone. The criteria for measuring enterprise emphasis include the frequency of occurrence and the contribution to total household labor and income. b. Examine extent to which income levels are a function of enterprise choice. The association between income levels and enterprise choice will be tested using comparative net margin analysis. Expected returns based on nationwide average enterprise returns are used to indicate the effect of enterprise emphasis on income level. I c. Present levels and timing of factor requirements and returns for major enterprises. Summarize effects of factor requirements on enterprise choice. An examination of factor requirements may help explain why farmers of different income levels choose different enterprises. One major hypothesis is that capital constraints prevent low-income farmers from pursuing high-returns enterprises. Whether poor house- holds select enterprises with low capital-labor ratios and high labor- land ratios is also investigated. The description of factor require- ments for individual enterprises, highlighted by an examination of seasonal labor use, provides background for meeting the objective which follows. d. For important enterprise combinations, discuss factor requirements and returns. Examine the degree of compatibility and conflict between enter- prises in combination. Summarize effects of factor requirements on choice of enterprise combination. This section focuses on how component enterprises of major enterprise combinations fit together with respect to factor require- ments. It is hypothesized that the enterprises in a combination are selected to equalize, as much as possible, the distribution of labor throughout the year. As mentioned previously, Bylerlee, et al., (1977) found that peak season labor demands in Sierra Leone constrain farm income. This study hypothesizes that the farmer seeks to overcome this constraint by choosing enterprises with complementary peak labor periods, rather than overlapping ones. The use of hired labor to release peak-period labor constraints is also examined. e. Discuss implications of analysis for agricultural development programs and policies affecting farming systems in Sierra Leone. A better understanding of factors affecting enterprise choice and choice of enterprise combinations can help project planners release constraints to increase production and modify farming sys- tems to increase rural incomes. Throughout the study, we will 10 highlight the implications of the above analysis on rural develOp- ment policies and proposals. 1.3 Source of Data The data examined in this study were collected as part of a comprehensive study of farm households in Sierra Leone during 1974/ 75 (Spencer and Byerlee, 1977). The project was conducted with funding by USAID, the Rockefeller Foundation, and Njala University College, Sierra Leone. A multi-phase stratified sampling design was used to select sample farm households for the study. The country was first divided into eight rural resource regions as shown in Figure 2.1. The ecolo- gical and demographic characteristics of each resource region are described in Table 1.1. The eight rural resource regions were then subdivided into enumeration areas used by the Central Statistics Office and three enumeration areas were selected to represent each rural resource region. Enumeration areas average about ten square miles in area and contain about 130 farm families. Sample frames of farm house- holds were then prepared for each selected enumeration area. From these frames, a stratified sample of twenty farm households and four nonfarm households (excluding traders) were selected at random in each enumeration area. The sample obtained is thus a regionally- balanced, representative sample of rural households. Between March 1974 and June 1975, selected households were visited twice weekly by resident enumerators to collect the 11 Figure 1.1 SIERRA LEONE RURAL RESOURCE REGIONSa 1 11' 11 i 1 1 C‘ “I __ f .................. " .1 .00 .II ’- -. “ht-4|. ."\.‘ C .I \. I J \\ i} "I. 7 “O“IC .\o 1 I ‘1 s ‘1" 6‘ r W“ \""o ' 9 L ‘\ (AZ-:1 I a I c : ’ 2 ' 7-1 5 .I' (i . i ".5.” a /. «hm: K. 0.”. ’:... . . fyRu‘IABpr .‘-" "'h\.\ ‘3 4“ folk-bl: but j! ' \ I 1 __ 0‘ c - \. . .. --" .- ”v \ (“mm ./ Mutant! :’ I “h 5 l \ oh’I Lot. (fin/3'3.” 1.3/7 I \o “In!“ ”I. 10.1 tutor :1. ”TV/'3“ ’ cum "/1 ":aflrw- 1‘ \.\ / 1.....3‘, _r .- 3..)- '. we -. ‘ o- .o .. ° ." _ . ,0 -. 02m 1.; .5. 3 S. "I ‘ 1 “rd"; .‘l } so .. .. .. «curs 'L / o. \,,A,./ f~-‘./ .. t. .1 ..-.-\ 5 < " \-. ." ) g o fie" ..\.‘ \o. LORI“. o. ', “Infant; \. 5‘ “M...“ ’.r,_,./ '. 8 ‘. Hobo: i’. f F- ") ’anli a. ' ,l. Sago-I" i — r : l.3 . ll ""fz'f "m" : \ r U‘. ..., ...... s. .2 5 . odorio..\ ‘. V. Pond: bu .’ k . AZ '1'? I .’ $ «- .0 F‘ .' I a (b . It. Hctuu/ ._ ) .I. \ m QfirAZ‘Hfi .. 7 .,.I \. : Hod-uh! Isn‘t-u" .. ’ I ) ,f' low-m Icgoml loo-Jory” ........_.q A? \-. a" ”+1.1 5K International Boundary ------ —---- no to so a so 00mm : " \ . L—J w: “LI-Ff" -'*:-'-—.—v_: “=4 /‘ (causation Ann-- . -- - - u ‘3 ‘h-‘i' 1' H o to a 0‘0 jun-mu ‘ NW“ "Ii.“ -------- 2 “V5 in“ 080 no ‘1’. '1. T" South North East Southern Coast (2) Scarcies (l) Moa Basin (6) Riverain Grasslands (4) Northern Pla1ns (3) Southern Plains (8) Boliland (5) Northern Plateau (7) aSpencer, Dunstan S.C. and Byerlee, Derek. 1977. "Small Farms in West Africa: A Descriptive Analysis of Employment, Incomes and Productivity in Sierra Leone." African Rural Economy Working Paper No. 19, Department of Agricultural Economics, Michigan State University, East Lansing,Michigan. 12 .wcmco .mczc c_ .oo;Um c:_scmuum mac cc mcmmx 9. cu m cwcu—wzo co acmucoac IIInllIIIIIIIIiIIII i.!tl|:||l:p_< mncm_mmcco some1oonm m. m.om o.om mm~.m muco— 1mmcco :wmcm>pm Pow>=—F< maswzm vcepc~ mascxm mesa» m_omo:u,4 o>ococwz swan m.om v.0 once: mu_cmucg acoucoumm coac1oomm o_. ~.~N o.em omv.m «meow ccwzuaom c.~m m.m 1 - 1 1 om_-o ~.om “.mko NN~.- nuzom excocox cap—3d mu_cwumg execs v.~_ m.o mas_4 gmwuvwm :mccm>em comm-ooom_ ocv1ooM 9.0— o.-~ won.om amoum—a :Lozucoz swan 0x04 xccncoomm mncmpmmeco mnE_4 esozm o.m. P.m m:Em» wu_cmac4 goccm>em Ommm-ooem omp-m_ m.~m ~.-_ v~o.v mvco—w_om swan xcovcoowm . zcccc>wm mmecc m.p_ m.~ mesa» muwcmucg ou_zac4 comm-Ocmm cm_-._ _.mv o.-~ oo~.m mews—a ccosucoz maEezm v:c_c_ magezm mu_cwuca ¢>ccccez swam czocn sop—ox wmmco ~.o~ N.m anew» cu gmwcucm E:_voacsmm;u Ommm-0cm~ m_. o.mo m.om~ «Fe.~ mmwucmom m.w_ w.o - - - . coe-o o.m~ m.cmw co_.am gucoz New I. - - - - 83 new 3%.. game was: 5am; Aucmuccov m~_m aaoco mrwom co_ucamco> figsv A_m>w_ mom m>onm A E; Aooo_v ANExV covmmz mwuccvc¢3u< v_o;mm:oz u_c;uu ._cc:mmm mcwumev gun mmomcmav :owumpzaoa cmc< possum moccw>< acmc_soo cam: cowac>wpm zummcwo acoewca covua_:goa pocaa “zen; \ a \ I : I -— z of total labor input goes to the particular enterprise. bHouseholds in which > 10: of the total labor input goes to the enterprise or > 10% of total income is generated by the enterprise. cHouseholds in which > 30% of total labor input goes to the enterprise or > 30% of total income is generated by the enterprise. dAn underestimate since northern mangrove swamps were not surveyed. n.a.: not available 26 households respectively. Ninety—three percent of the households pursue nonfarm occupations with fishing being undertaken by 40 per- cent, small industries by 31 percent, and labor sold out by 69 per- cent of the households. Column 5 shows the percentage of households in which an enterprise is important. An enterprise is important for a household if it contributes greater than 10 percent to total income or labor. Upland rice is important for over 75 percent of all households and wild oil palm is important for about 40 percent. Cassava, inland swamp rice, labor sold out, and groundnuts are other enterprises which are important for over 20 percent of the households. According to Table 3.1, a household specializes in an enterprise if the enterprise accounts for more than 30 percent of total labor or income. Seventy-one percent of all house- holds specialize in upland rice and 20 percent specialize in wild oil palm. Only four other enterprises, inland swamp rice, cassava, onions-peppers-tomatoes, and labor sold out, are specialized in by more than 5 percent of the households. The average household special- zes in 1.5 enterprises. The percentage contribution of different enterprises to total labor, income, and acreage of households is shown in the last three columns of Table 3.1. Upland rice accounts for 47 percent of total labor and 27 percent of total income, leading both categories. Inland swamp rice, wild oil palm and labor sold out each account for over 5 percent of total labor and income. In addition, 27 groundnuts accounts for over 5 percent of total labor and fishing accounts for over 5 percent of total income. Rice accounts for over 80 percent of total acreage cultivated; other annuals and tree crops represent 11 percent and 7 percent, respectively. Table 3.1 also shows the contributions of nonfarm enterprises to household income and labor. About 21 percent of total household income in rural Sierra Leone comes from nonfarm sources. Nonfarm labor accounts for 13 percent of total labor used by the rural house- hold. Two enterprises, labor sold out and fishing, account for over half of nonfarm labor and income. 3.2 Enterprise Returns Tables 3.2 and 3.3 present net enterprise returns per house- hold man-hour equivalent1 and per acre, respectively, for eighteen enterprises. The returns are computed by calculating the value of output for an enterprise and subtracting actual or imputed values for variable costs (hired labor, land payments, seed, fertilizer, and mechanical services, etc.), an establishment cost factor (for tree crops), and an annual cost of capital factor (especially important for nonfarm enterprises). Net returns are then divided by the number of household man-hour equivalents or acres devoted by the household to the given enterprise to arrive at enterprise net returns per 1Female and child labor hours are weighted at .75 and .50 man-hour equivalents respectively, reflecting the ratios of their hired labor wages to those for adult males. Adults are those house- hold members 16 years and older (Spencer and Byerlee, 1975). 28 Table 3-2 FARM AND NON-FARM ENTERPRISES CLASSIFIED ACCORDING TO NET RETURNS PER MANHOURa Returns to Labor By Region (cents/hour) Returns Category Number of and Enterprise North South East National Observations eve Fundi 5.4 - - 5.4 33 Labor Sold Out 5.9 7.8 7 5 6.9 228 Upland rice 6.9 7.7 10.8 7.9 227 Groundnuts 12.2 5.9 - 9.9 62 Onions-peppers-tomatoes 10.0 - - 10.0 25 Middle Carpentry b - - - 12.1 16 Inland Swamp Rice 11.1 15.8 15.8 12.5 46 Coffee - - 16.8 16.8 27 Cassava - 23.7 - 19.9 79 Riverain rice (mech.) - 23.8 - 23.8 12 Oil Palm (wild) 16.0 28.1 44.8 25.4 120 Hi h Blacksmithing - - - 27.7 14 Mangrove rice - - - 27.9 11 Tailoring - - - 32.1 19 Cocoa - - 33 5 33.5 13 Boliland rice (mech.) 35.7 - - 35.7 9 36.8 - - 36.8 13 quly households for which an enterprise accounted for more than 10 per cent of total household labor or income are included in the computation of net returns for that enterprise (exception is labor sold out for which all households selling labor are included). Blanks are shown above where there were less than 10 households in the given region meeting the above criteria. bFigures for the South and East have been combined due to an insufficient number of cases for each region individually. Source: Survey Data 29 manhour: or per acre. Returns data for an enterprise are based on households in which the enterprise contributes greater than 10 per- cent of total labor or income (Spencer, Byerlee, and Franzel, 1979). EnterpriseS'hiTable 3.2 are grouped into high-returns, middle- returns, and low-returns per amnhour enterprises. Inland fishing has the highest return, 64.7 cents per nmnhour. Marine fishing, mechanized Boliland rice, cocoa, tailoring, mangrove rice, and black- smithing follow. The enterprise with the lowest returns is fundi, 5.4 cents per hour. Other low-returns enterprises include labor sold out, upland rice, onions-peppers-tomatoes and groundnuts. Middle- returns enterprises include wild oil palm, mechanized riverrain rice, cassava, coffee, inalnd swamp rice, and carpentry. Table 3.2 shows the high returns associated with nonfarm enterprises. Of the six nonfarm enterprises listed, four are high- returns enterprises. Although nonfarm enterprises make up only one- third of all enterprises, they account for 57 percent of the high- returns enterprises. An inter-regional comparison shows returns per manhour to be consistently lower in the North than in the rest of the country. The North has the lowest returns per hour of any region for upland rice, inland swamp rice, oil palm, and labor sold out, enterprises which account for over half of total income. The North's lower enterprise returns are associated with the region's poorer ecological character- istics and contribute to the lower incomes per consumer equivalent recorded in that region. For the three major enterprises which are 3O pursued in both the South and the East, the East has higher returns for two of them--up1and rice and oil palm. Net returns per acre for major crops are shown in Table 3.3. Onions-pepper-tomatoes have the highest returns to land, 366 Le. per acre, or about three times as high as mangrove swamp rice, the next leading crop. acre. Fundi and upland rice have the lowest returns per TABLE 3.3.--Net Returns per Acre for Major Crops in Sierra Leonea Enterprise Net Returns Per Acre gggber if South North East Nationwide erva ions Rice Upland 35.73 34.45 54.11 37.00 227 Inland Swamp 100.68 97.26 100.68 98.52 46 Mangrove 121.78 --- --- 121.78 11 Riverrain 55.08 --- --— 55.08 12 (mechanized) Boliland --- 54.60 --- 54.60 9 (mechanized) Other Annuals Fundi --- 34.55 --- 34.55 33 Cassava 56.18 --- --- 56.18 79 Groundnuts 40.81 68.81 --- 58.73 62 Onions-Peppers- Tomatoes --- 353.04 --- 353.04 '25 Tree Crops Cocoa --- --- 65.69 65.69 13 Coffee --- --- 70.74 70.74 27 aBlanks indicate that less than ten observations were avail- Figures for inland swamp ricein the South and East are combined due to a shortage of observations in the individual able for analysis. regions. 31 3.3 Differences in Enterprise Emphasis Among Income Groups The households in this study are assembled into income groups according to their incomes per consumer equivalent (ICE). ICE is employed instead of per capita income to account for differences in consumption requirements among families with different sex and age structures. This is important because of the high degree of varia- tion in the composition of households among different regions of Sierra Leone. In general, households are larger and dependency ratios higher in the North. Family members are accorded consumption weights as shown in Table 3.4. These weights were established by the FAO for the TABLE 3.4.--Age/Sex weights Used in Calculating Consumer Equivalents Age/Sex Male Female O-4 .2 5-9 . .5 10-14 .75 .7 15+ 1.0 .9 SOURCE: FAO, 1957. calculation of man-equivalent calorie requirements (FAO, 1957). Since food consumption accounts for about 70 percent of total con- sumption in rural Sierra Leone these weights are believed to be reasonable proxies for overall consumption requirements (King and Byerlee, 1977). 32 The sample households are divided into income terciles with 98 households (30 percent) in the highest and lowest terciles, and 132 households (40 percent) in the middle tercile. They are also divided into income deciles with 32 to 33 households (10 percent) per decile. The analysis of enterprise emphasis among income groups is important for designing programs to increase the incomes of the rural poor. In order to help low income farmers, we must understand which enterprises they pursue and the constraints which limit their particular production possibilities. Table 3.5 shows that enterprise emphasis changes significantly between income groups. Twelve of the 22 most important enterprises are associated with a specific income tercile, using a test of proportions and a 10 percent significance level (Clark and Schkade, 1974). Nine enterprises are significant at the 5 percent level. Boliland rice (hand and mechanized) and blacksmithing are emphasized by high-income households, whereas groundnuts, onions-peppers-tomatoes, marine fishing, and carpentry are undertaken primarily by middle-income households. Low-income farmers emphasize upland rice, fundi, other vegetables, fruits, and labor sold out more often than other income groups. A comparison of the results of Table 3.2 and Table 3.5 show that all enterprises emphasized by high-income households have high returns to labor, and that all enterprises emphasized by low-income households have low returns to labor. The enterprises emphasized by middle-income households include low, middle, and high-returns enter- prises. 133 TABLE 3.5.--Enterprise Emphasis Among Income Group in Rural Sierra Leone Percentage of Households for Which Income Class Enterprise Contributes Greater than Number of Emphasisa 10% of Total Labor or Income Observations Level of High Income Middle Income Low Income Significance .05 .1O FARM 31s: Upland 69.4 76.5 88.8 25.6 -- L Inland Swamp 28.6 27.3 24.5 88 -- -- Mangrove 3.1 1.0 O 4 Boliland Rice (Hand) 7.1 4.5 1.0 14 -- H Boliland Rice (Mech.) 6.1 2.3 0 9 H H Riverain (Mech.) 4.1 4.5 3.1 13 -- -- Other Annuals Fundi 7.1 9.8 16.3 36 L L Cassava 28.6 28.8 25.5 91 -- -- Groundnuts 17.3 28.8 19.4 74 M M Onions-Peppers-Tomatoes 8.2 12.1 4.1 28 M M Other vegetables 4.1 10.6 16.3 34 L L Tree Crops Fruits O 2.3 5.1 8 L L Cocoa 8.2 5.3 5.1 20 -- -- Coffee 11.2 9.8 12.2 36 -- -- Oil Palm (wild) 36.7 40.1 43.9 132 -— -- Animals 0 0 2.0 2 NONFARM Fishing Marine Fishing 6.1 8.3 2.0 19 -- M Inland Fishing 7.0 8.3 3.1 21 -- -- Hunting and Gathering 2.0 2.3 2.0 7 Small-Scale Industries Tailoring 6.1 4.5 4.1 16 -- -- Carpentry O 4.5 2.1 8 M M Blacksmithing 8.2 3.8 O 13 H H Spinning-Weaving 1.0 1.5 O 3 Other small industries 3.0 3.8 5.1 13 -- -- Trading 2.0 3.8 1.0 8 -- -- Labor Sold Out 18.4 20.5 39.8 84 L L aUsing Test of Proportions(Clark and Schkade, 1974). H = High; M = Middle; L = Low. eight cases are tested. Only those enterprises with over 34 Therefore, it is likely that enterprise choice is an important factor in determining rural income levels. Since poor rural households pursue low-returns enterprises, they obtain low levels of income. Table 3.6 compares enterprise emphasis among the high decile and low-decile households. The table shows the heavy emphasis which the poorest households give to upland rice. Upland rice labor accounts for 67 percent of total labor inputs for low-decile house- holds but only 30 percent for high-decile households. Households in the high-decile group devote a greater proportion of their labor to virtually all other enterprises (exception being fundi and carpentry) than do the poorest households. The differences in enterprise emphasis between high and low- decile groups is somewhat different than the differences between high and low-tercile groups. Upland rice, fundi, and "other vege- tables" are emphasized by low-decile households. Inland swamp rice, cassava, onions-peppers-tomatoes, fishing, and tailoring are under- taken primarily by high-decile households. Three less prevalent enterprises, mechanized riverain rice, mangrove swamp rice, and cocoa also make important contributions to the incomes of high-decile households. 3.4 Comparative Expected Net Mamin Analysis: Returns to Labor and Land The preceding analysis shows that enterprise choice varies 1 among income groups. Next, expected net margin analysis is used 1Upton refers to this method as potential net margin analysis. 35 TABLE 3.6.--Enturprise Emphasis Among the Highest and LOwest Decile Income GrOups in Rural Sierra Leonea 'r. l"‘..1‘r!' ‘3‘. 1|»:": 911.: 3'11: 1 I. x 4.1:“ r - ; are i .31 Percentage of Heuseholds Undertaking Enterpriseb Percentage of Heuseholds for which Enterprise is :r : 15:3.1 Ar‘fimm1.-—“t “'1 .3 fl rm"; 3 asl‘a‘ Percent of Total Percent of Total Heusehold Labor tar-$1.81.; v -rt';.; H0usehold Income ImportantC - High 10‘ Low 10 High 10 LOw 10. High 10" Low 10. High 10 Low 10' £58.! Lice Upland 66 91" 14 28 30.4 66.7 18.7 32.8 Inland Swamp 53" 37 19 9 9.8 8.0 11.9 4.1 Mangrove 3 0 3 O 3.0 O 3.4 O Boliland (hand) 9 3 3 O O 0 0.1 O Boliland (mech.) 6 O 3 O 2.5 0 6.3" O Riverrain (mech.) 9 3 9 3 3 1.2 7.1“ O Other Annuals Fundi 6 31" 3 3 1.4 4.6 0.4 2.0 Cassava 75“ 41 3 6 4.9 1.9 3.1 7.2 Groundnuts 47 47 6 0 5.1 2.8 2.9 4.3 Onions-Peppers-Towatoes 22“ 6 9 3 6.3 0.7 4.9 5.1 Other Vegetables 34 53“ O O 0.6 0.6 0.2 0.7 L166 Ewes Fruits 19 9 0 3 0.6 0.2 -- '- Cocoa 19 9 3 O 1.4 0.2 4.8:“ 0 Coffee 25 16 3 O 1.8 1.0 4.4 2.0 Oil Palm (Hild) 59 66 16 25 5.6 3.4 6.1 16.7* Animals 3 6 0 3 0.1 0.2 0.1 1.0 NONFARM EIEPIDS 50" 16 9 O 4.3 0.3 16.5*' 0 MOEQDB 22 22 o 0.3 0.2 0.5 1.6 Small Scale Industries Tailoring 22“ 3 9 3 2.6 0.2 4.6 3.0 Carpentry 3 6 O 3 0.2 0.9 0 0.5 Blacksmithing 6 12 3 O 0.3 0 0.8 O Spinning-weaving 9 3 O 0 0.3 0.1 0.5 O Other Small Industries 12 2 O 3 2.3 0.1 1.4 0 Trading 6 6 O 3 0.2 0.2 0.3 1.6 Labor Sold Out 72 72 O 9 5.8 5.3 2.4'” 17.1"“ TOTALS 100.0 100.0 100.0 100.0 a, and '* means significant at the 106 and 5’ level respectively under a binomial distribution test for small sample sizes. i and ti mean significant at the 10% and 5: levels respectively, under a Chi-square test. bEnterprise contributes greater than 1% of total household labor or income. cEnterprise contributes greater than 10E of total household labor or income. dwhere returns to an enterprise are negative, zero is substituted for a negative percentage. 36 to measure the effect of enterprise choice on labor productivity and land productivity (Upton, 1973). Since labor is the chief resource constraint to increasing incomes in rural areas, labor productivity, as measured by the returns to labor, is emphasized (Spencer and Byerlee,l977). The expected net margin per hour for an enterprise is the average returns per hour for that enterprise for all farmers under- taking it. The expected net margin, thus, gives an expected value of returns based on the performance of all farmers engaged in the enterprise. The overall household expected net margin per hour weights the various expected net enterprise margins per hour by the amount of labor expended for each enterprise. It thus represents the returns to labor based solely on the choice of enterprises and not on the actual performance of the particular household. A numerical example is provided for further clarification. Assume that two farmers grow only two crops each-~groundnuts and cassava. Farmer A devotes 80 percent of his labor to groundnuts and 20 percent to cassava while Farmer B devotes 50 percent of his labor to each cr0p. Looking back at Table 3.2 we see that the nationwide net returns per manhour for groundnuts and cassava are about 10 cents and 20 cents per hour. The farmers' individual expected net margins per hour are then weighted by percentage of total labor expended: Farmer A: (.8 x 10) + (.2 x 20) 12 cents/manhour Farmer B: (.5 x 10) + (.5 x 10) 10 cents/manhour 37 We thus conclude that Farmer A pursues an enterprise mix which gives him a higher expected net margin per hour than Farmer 8. Expected net margins to land are calculated in a similar manner. But whereas expected net margins per hour are weighted by the enterprisefispercentage of total labor, expected net margins per acre are weighted by the enterprisekspercentage of total household land. Net margins per acre are obtained from Table 3.3. As stated previously, studies by Petu and Upton (1964) and Matlon (1977) showed that the crop mix accounted for very little variation in productivity. Matlon, however, found a greater emphasis on nonfarm enterprises among high-income households which did result in higher overall returns to labor. In the following analysis, expected net margins per hour are shown for both farm enterprises and all enterprises undertaken by the households. A comparison of expected net margins per manhour for farm enterprises of different regions is shown in Table 3.7. In column 1, expected enterprise net margins per manhour are based on the regional averages of enterprise net margins from Table 3.2. In column 2, net margins per manhour are computed using nationwide averages. When general averages are used, the net margin per hour takes into account differing productivity between acres for a par- ticular enterprise. National averages, on the»otherhand, gloss over regional differences in productivity. In both cases, the differences between mean expected net margins of different regions are signifi- cant at the 1 percent level, using an analysis of variance test. 38 TABLE 3.7.--Expected Household Net Margins to Labor and Land Among Regions of Rural Sierra Leonea Returns to Labor Mean Expected Household Net Margin/Manhour (cents) Using regional Using Nation- Returns to Land Mean Expected Household Net Margin/Acre (Le.) Using Regionwide $23.32?“ ”SliieECEEEins Enterprise Returns Farm Enterprises Only South 12.96** 13.63** 45.04 North 9.93** 10.69** 79.31 East 14.91** 12.33** 68.77 National 11.91 12.20 62.56 All Enterprises South 13.44** 13.97** -- North ll.01** 11.77** -- East 15.21** 12.80** -- National 12.69 12.87 -- aExpected household net margins are defined in Section 3.4. Enterprise returns to land and labor are obtained from Tables 3.3 and 3.4 respectively. Where there were less than ten cases of an enterprise in a region, nationwide net margins for the enterprise were used. *Differencesbetween regions are significant at the 5 percent level using a test of analysis of variance. **Differencesbetween regions is significant at the 1 percent level, using a test of analysis of variance. 39 Using nationwide average returns, the South has the highest expected net margin per manhour, 13.7 cents, whereas the North has the lowest 10.7 cents per manhour. The predominance of high-returns enterprises in the South, e.g., mangrove rice, mechanized riverain rice, and inland fishing, explains the South's high ranking. Crops found in the North, e.g., fundi, groundnuts, and onions-peppers-tomatoes, tend to give low returns. Assigning regional averages for net margins per manhour to households within each region (column 3) results in a different ranking. The East is highest at 14.9 cents per hour, since returns to labor for major enterprises are higher in the East than in other regions. Both the crop mix and differing returns of major enter- prises between regions leads to inter-regional differences in labor productivity and income. Table 3.8, column 1, presents the expected farm net margin per manhour for different income groups. Nationwide, the high- decile group has an expected net margin 40 percent higher than the low-decile group and this difference is significant at the 5 percent level. Differences between the terciles were not significant. Nevertheless, from the highest income group through the lowest income group in Table 3.8, there is a consistent trend of decreasing expected household net margins per hour for farm enterprises. Therefore, there is some degree of association between crop mix and income level. 40 TABLE 3.8.--EXpected Household Net Margins to Labor and Land by Region and Income Group in Rural Sierra Leonea Returns to Labor Mean Expected Household Net Margin/manhour (cents) Returns to Land Mean Expected Income Group Household Farm A11 Net Margin/Manhour Enterprises Enterprises (Le.) South 12.96 13.44* 45.04 Lowest 10% 13.61 12.20** —- Lowest 30% 12.80 12.89* 41.72 Middle 40% 12.28 12.71* 48.12 Highest 30% 14.05 15.00* -- Highest 10% 19.34 18.52** 44.17 North 9.93 11.01 79.31 Lowest 10% 9.64 8.25* -- Lowest 30% 8.71 8.72** 76.83 Middle 40% 11.01 ll.36** 80.18 Highest 30% 9.70 12.83** 80.59 Highest 10% 10.30 12.47* -- East 14.91 15.21 68.77 Lowest 30% 14.84 15.23 47.50 Middle 40% 14.77 14.96 89.04 Highest 30% 15.17 15.50 64.56 Nationwide 12.20 12.87** 62.56 Lowest 10% 10.73* lO.80** -- Lowest 30% 11.34 ll.35** 61.60 Middle 40% 12.45 13.24** 63.51 Highest 30% 12.71 13.90** 62.26 Highest 10% 15.06* l6.51** -- aExpected household net margins are defined in Section 3.4. Enterprise returns to land and labor are obtained from Tables 3.3 and 3.4 respectively. Where there were less than ten cases of an enterprise in a region, nationwide net margins were used. Decile data are not shown for the East because deciles in the East consisted of less than ten households. *Difference between regions is significant at 5 percent level using a test of analysis of variance. **Difference between regions is significant at the 1 percent level, using a test of analysis of variance. 41 Table 3.8, column 1, also presents a breakdown by income group for each region. In this analysis, there is less indication of an important association between income levels and expected farm net margin per hour. At a 5 percent level of significance under a test of analysis of variance, there were no significant differences in expected farm net margin per hour between either tercile groups or high- and low-decile groups in any of the regions. Nor in any region did expected net margins follow a consistent declining trend from high-income groups to low-income groups. Tables 3.8 also shows expected net margins per hour for all enterprises, that is, both farm and nonfarm activities. The results are nearly identical to those in the analysis of farm enterprises, with differences between regions significant at the 1 percent level. Therefore, the enterprise mix is important in generating differences in productivity between regions. Table 3.8, column 2, presents all-enterprise expected net margins per hour for different income groups. The high-tercile households have an expected net margin per hour 5 percent higher than that of the middle-tercile and 22 percent higher than that of the low- tercile households. Moreover, the differences between income groups are significant at the 1 percent level. Thus, a strong association exists between income level and expected household net margin per hour. This supports the hypothesis that enterprise emphasis has a significant effect on income distribution. 42 Similar results are obtained at the regional level. In both the South and the North differences between high- and low-decile groups are significant at the 5 percent level, under a test of analysis of variance. In the North, differences between terciles are significant at the 1 percent level and follow a consistently decreasing trend. Although the low-tercile group has a slightly higher expected net margin per hour than the middle-tercile group in the South, when they are grouped together they are significantly lower than the high-tercile at the 1 percent level. In the East, differences between income groups are not signficant. Several important points can be made by comparing the all- enterprise figures with those for farm enterprises only. Using nationwide net margins per hour, the nationwide expected net margin per hour for all enterprises is 12.9 cents, 6 percent higher than that for farm enterprises. Furthermore, Table 3.7 shows that the disparity between the high-returns region and the low-returns region is 25 percent greater in the farm-enterprises analysis. Thus, non- farm enterprises both increase the net margins per hour and con- tribute to equalizing the expected net margins per hour between different regions. Nonfarm enterprises, however, appear to skew the nationwide distribution of income. The range in expected net margins per hour between terciles is 12 percent for farm enterprises and 22 percent for all enterprises. Corresponding figures for the ranges between high— and low-deciles are 22 percent and 53 percent respectively. 43 Nevertheless, differences in expected household net margins among both regions and income groups are small when compared with actual differences in incomes per consumer equivalent. ICE differ- ences between the East (the region with the highest ICE) and the North (the region with the lowest ICE) were 30 percent whereas expected net margins differed by only 19 percent. Differences in expected net margins play an even smaller role in determining income differences between income groups. The high-tercile group's ICE is 5.7 times that of the low-tercile group whereas its expected net margin per hour is only 22 percent higher. ICES in the high-decile group are 14.9 times those of the low-decile group; expected net margins per hour differ by only 53 percent. In conclusion, significant differences exist between the expected net margins per hour of income groups and regions, and nonfarm enterprises are especially important in determining these differences. However, the differences in expected net margins between income groups are small when compared to actual differences in income levels. Thus, enterprise choice has a limited role in determining the distribution of income. Tables 3.7 and 3.8 also explore the relationship between enterprise choice and retuns to land. Table 3.7, column 3, shows expected net margins per acre by region. The differences between the regions are not significant at the 5 percent level. Nor is there a significant relationship between expected net margins per acre and income levels at the nationwide and regionwide levels (Table 3.8, 44 column 3). Nationwide, the expected net margin per acre differs by only 3 percent between the income terciles with the highest and lowest values. Thus there is no tendency for high-income farmers to choose crops which have higher overall returns per acre. This finding supports the hypothesis that access to land is not a major con- straint to increasing incomes. 3.5 Enterprise Combinations A study of enterprises in combination with their respective factor requirements can provide a better understanding of the farm system. For the purposes of this study, an enterprise combination consists of two enterprises which account for at least 60 percent of a household's total labor'andwhich individually account for at least 10 percent of total household income or labor.1 An enterprise combina- tion serves as a simplified approximation of an overall farm system. Nine enterprise combinations,each representing at least ten house- holds,are shown in Table 3.9. All combinations include upland rice as one of their component enterprises. Each of the nine enterprise 1According to this definition, one household may have more than one enterprise combination. In the initial stages of this study, it was decided that enterprise combinations would be defined in terms of households, i.e., one enterprise combination per house- hold. A household's enterprise combination was defined as all enter- prises contributing greater than 10 percent of total household labor and/or income. However, because of the great diversity in enter- prises and enterprise combinations, few households had the same com- bination. Using a sample of 328 households, there were only two enterprise combinations (upland rice alone and upland rice-wild oil palm) which represented over 12 households. Different values of the percentage contribution to total labor and income were tried, to no avai . .uaauso 0cm: E.wa .mmmmeaemuzm mcwgm.»-mewems 0:» van ucm.cm m. moua.uc. v oc.e:mcme mo xu_..ncw.mcca me. o. man .vmva.uxm use zucoz mg. :. memes comuuewezcw 03¢ :. momwcacoucm spam ._oU .No. u moccuwm_cm_m we .m>m— .mcowugoaoga No mummy ou mcvvgouu88 8.882 8;. 8. 08880 88.88888888 83. :. 088.888.8588 2.88 ..o188.8 n88.828 .m.m co..umm :. n88.88n 888 088.888.8588 80.88.80800 8.0N 8..8 N.0 ..8. 8.88 0.0. 888888888..88 nac..0..0. 88.0 888.88 0.0. ..88 ..8. 8.00 8.80 N.N. .8888 88.0 888.88 ..0. 8.00 ..N 8.8. 8.00 0... 808.80.. 88.0 888.88 8.0. 8.00 .... ...8 8.00 0.N. 888.88 88.0 888.88 8... 0.00 0.0. 8.80 N.80 0.0. 88.0 888:0 888.8. 88.0 888.88 0.0N 8.0N 0.0 8.08 0.80 0.0. 8880088 88.0 888.88 ..8. 8.08 ..0. 8.88 0.08 8.0. 0888888880 88.0 888.88 8.0. 8.80 N.8. 0.00 8.00 8... 888 8.80 88888 88.0 888.88 0.88 ..80 8.0. ..08 0.8. 8.0. ”8.88 ..8 8.0 888.88 8888:. 8888:. n.88 n.o;8088: 8088. 8888:. 88888 -8088: .888. .888. 88 88888 .888. .888. 88 8.8880888 .888. 88 88 88.888 88.888.88 88 88.888 88.888.88 88 88.888.88 88.888.88 1.8.880 x :88: .889 N :88: -.88880 0 :88: 1:00 0 :88: -cou x :88: 1:00 x 8882 80.8888880 ncm 88.0 n88.88 80.88.8880 new 88.0 n88.82 88.888.8880 88.808.8580 88:88. 8888.0 .8880 8. 8888:. new 8888. n.ax -8088: .888. o. 088.888.8880 80.8888880 :. 080.8888880 888888580 .8 088.888.8ucouui.o..m mam<. 48 Upland rice accounts for over half of total labor for all combina- tions. The contributions of the second enterprise to total labor range from 3 percent (fishing) to 20 percent (inland swamp rice). Contributions of upland rice returns to total income range from 25 percent (in combination with cassava) to 49 percent (in combination with groundnuts). Second-enterprise contributions range from l0 percent (groundnuts) to 41 percent (wild oil palm). Zhi two enter- prise combinations, upland rice with cassava and with wild oil palm, upland rice contributes less than the second enterprise to total income. The association between enterprise combinations and income levels for different regions is explored in Table 3.ll. The South and the East, which have similar physical characteristics, are con- sidered to be a single region in this analysis. Upland rice-wild oil palm, is emphasized by low-income farmers in the North but shows no marked association with any income group in the East-South. In the nationwide analysis (Table 3.9) it is associated with low-income households. Upland rice-inland swamp rice, which is not associated with a particular income group in the nationwide analysis, is clearly associated with low-income farmers in the North and with high- and middle-income farmers in the East-South. Upland rice-labor sold out is most commonly found among low income farmers in both regions, which mirrors the results of the nationwide analysis. Upland rice-- groundnuts, which is emphasized by middle-low income groups in the nationwide analysis, is not associated with any income group in either 49 TABLE 3.ll.-~Enterprise Combinations and Income Levels of Rural House- Holds in Regions of Sierra Leonea No. of Cases Rer Income Group Enterprise Combination Region Cases Income High Middle Low Levelb (30%) (40%) (30%) Upland Rice--wi1d N 11 O 2 9 Low Oil Palmc (18%) (82%) Upland Rice--Wild Oil E-S 58 14 25 19 -— Palm (24%) (43%) (33%) Upland Rice--Labor N 12 1 2 9 Low Sold Out (8%) (17%) (75%) Upland Rice-~Labor E-S 36 8 13 15 Low Sold Out (22%) (36%) (42%) Upland Rice-~In1and N ll 2 2 7 Low Swamp Rice (18%) (18%) (64%) Upland Rice-~In1and E-S 29 12 12 5 High- Swamp Rice (41%) (41%) (17%) Middle Upland Rice--Ground- N 18 4 9 5 -- Nuts (23%) (50%) (2 %) Upland Rice-~Ground- E-S 26 5 l3 8 Nuts (19%) (50%) (31%) aEnterprise combinations are defined in Section 3.5. Only those enterprise combinations with more than ten cases in the East- South and North are included. bAccording to test of proportions (a .10). cOil palm enterprises in two enumeration areas in the north have been excluded, due to the unreliability of measuring palm output. 50 region. This analysis thusshows that relationships between enter- prise combinations and income groups which occur nationwide do not necessarily hold at the regional level. Both the upland rice-wild oil palm and the upland rice- inland swamp rice combinations are associated with poorer farmers in the North than in the East-South. This result can largely be attri- buted to the lower returns per man-hour for each of the three crops in the North as compared to the other two regions. This, in turn, reflects the poorer physical characteristics of the North. The association between enterprise combinations and income levels will be further examined in Section 5. CHAPTER IV FACTOR USE AND ENTERPRISE EMPHASIS This chapter examines the levels and timing of factor use and the relationship between factor use and enterprise choice. Emphasis is given to the use of labor and capital, the two most severely limiting factors in rural Sierra Leone. The complimentarity of upland rice and inland swamp rice with other enterprises is also examined. The analysis in this chapter provides background for the examination of factor use in enterprise combinations in Chapter V. 4.1 Seasonal Labor Use Among Regions The monthly distribution of labor used by households of dif- ferent regions is shown in Figure 4.1. Peak periods and slack periods in the South are relatively uniform, with the former extending from June through November and the latter from December through May. These periods correspond rather closely to the rainy season and dry season, respectively. The low coefficient of variation, 17.6, reflects the relative smoothness in labor use from month to month. Labor use in the north, on the other hand, is characterized by much sharper peak and trough periods. Peak months are July and August and the slack period is December through April. The coefficient of variation between months is 31.1, almost double that of the South. The East's peak 51 m maximum mm WHILE"! 52 Figure 4.l Monthly Distribution of Labor Used Per House— hold By Region in Rural Sierra Leone, May 1974-April, 1975 A. SOUTH B. NORTH I” no .. ion “96” .. “2"— § m r-l—_ ...7 3% d—r .‘35 mm 355 306 a ‘00 3 .0 200 ion IJJAsonoJruA III-am C. EAST D. NATIONAL no an M E .8 :3: In ZN i .. 53 labor period is June to September with a relatively uniform slack period extending through the rest of the year. The region's high monthly coefficient of variation, 27.0, is probably related to the 1 The high degree of homogeneity of cropping systems in the East. more seasonal nature of labor use in the North is related to its shorter rainy season and lower levels of rainfall. 4.2 Seasonal Labor Use for Individual Enterprises The monthly distribution of labor used for major enterprises is shown in Figures 4.2 to 4.10. Monthly labor profiles are on a per-acre basis for farm enterprises and a per-case basis for nonfarm enterprises. The analysis includes only those cases in which the enterprise contributes greater than 10 percent of total household labor or income. The peak periods and slack periods of individual enterprises are shown in Table 4.1. Labor profiles for upland rice are shown in Figure 4.2. Peak periods are June to November in the South, June to October in the North, and June to July in the East. In the North and the South, these periods correspond roughly with the rainy season. In the South and the East, the peak month for labor use is July, when weeding takes place. In the North, October, the month of harvesting, is the busiest month although labor use is high in July as well. lIt is recalled that the South and North consist of three and four resource regions, respectively, whereas the East consists of only one. A higher degree of variation in physical resources brings about more crop diversity and a consequent lower coefficient of varia- tion for seasonal labor use. Hm WINE!” 54 Figure 4.2 Monthly Distribution of Labor Used for the Production of An Acre of Upland Rice in Sierra Leone May 1974-April 1975 A. SOUTH B. NORTH 1W 1” F f: 90 no r“- .0 TL —)i so rm- iFs-FGLFI- g ”.32. .- § ‘0 33 3‘ E ‘0 -1 i 29 to 3' Zl 20 ,6 a 0 7 6 8 o n J J A 3”“; I o J I II A II J J A s o l o J I ma hath C. EAST D. NATIONAL flnfiu E ” n rt 77 g .4. r“, (IL 72 g 5 L:- ed 3” T‘ i” ! JL, g oz .0 T‘- 3.08%. 31 g 27 18 u .‘I c 0 A] uaaasouoarnu "J‘SOIDJFIA m mums Figure 4-3 55 Monthly Distribution of Labor Used for the Production of an Acre of Inland Valley Swamp Rice and an Acre of Mangrove Swamp Rice in Sierra Leone, A. May 1974-April l975 INLAND SHAMPS: 86 T SOUTH & EAST l3 naaasouoarna mm C. INLAND SWAMPS: NATIONAL m m ...I— .8 E— in ns __ m .. l— .“ ’2. L 56 0 :LF' __ an 1: ll? 0 uaaasonoaruia mm m mums “mum: B. INLAND SNAMPS: NORTH 1'60 20' m .. FF ISO "5 120 i N El f‘ r- .0 3|- 4! N I.— 30 0 L115 "In] I J a A s o a 0 air a . M“ D. MANGROVE SHAMPS: SOUTH '°° r“ I“ 110 1” '° '01 FL '° Ti“ 15 ’ O Inn-m WWII“! 56 Figure 4.4 Monthly Distribution of Labor Used for the Production of an Acre of Riverain Rice and an Acre of Boliland Rice in Sierra Leone, May l974-April l975 A. RIVERAIN-MECHANIZED B. BOLILAND-MECHANIZED w 2.. w 9 AL & 4L #L 5 a § 2:. a 27 3 n '8'" a 3L *5L3;' a m a a II 12 w 6 9 w s 2 Z 2 I" 0 r1: I J J I S 0 I o J F H i I 7 J J ‘ 5’ 0 I D J F H A ”I"! M" C BOLILAND-HAND n 4L .3- n 59 a I: 10 ‘lot‘ IJJISOIDJI’II I” 320 MM mums I” I NO _ A. FUNDI 57 Figure .4.5 Monthly Distribution of Labor Used for the Production of an Acre of Fundi and an Acre of Groundnuts in Sierra Leone, May 1974-April 1975 Ln 0 B. GROUNDNUTS: :2 ISO LL33 g m a d E .. IL so .9. 30 ..IL I J J A C. GROUNDNUTS: 50' In“ SOUTH J'IA IJJASOI NORTH D. GROUNDNUTS: NATIONAL nfi ADJ-'1 5 ”1 £- 9 ‘° 53 I .. lo I 0 2r—ll naensonoat A 58 Figure 4.6 Monthly Distribution of Labor Used for the Production of an Acre of Cassava and an Acre of Onions, Peppers, and Tomatoes (OPT) in Sierra Leone, May 1974-April 1975 A. CASSAVA: SOUTH M4" comm B. ONIONS, PEPPERS AND TOMATOES: NORTH In meL WW”?! I” ,0 59 Figure 4.7 Monthly Distribution of Labor Used for the Production of an Acre of Coffee and an Acre of Cocoa in Eastern Sierra Leone, May l974-April 1975 A. COFFEE .. “l ... I 5160 Em : AL lo 72 ‘0 a, 2 O nJJl$°'°"'"‘ "It“ 8 COCOA ‘20 IN Eco '2'... .5‘2. 3.. m 1 1° 1 n 9 5‘5 nun-m [WIVALEITS A. SOUTH 62- ‘7 VV . 3’ l H mm (Wl'fllm in \VS‘ *- 60 Figure 4.8 Monthly Distribution of Labor Used for the Production and Processing of Wild Oil Palm Products in Sierra Leone May l974-April 1975 B. NORTH 7///, tumult. a main. ,2]. no 3}. .32. too E g 80 E g .. C. NATIONAL a. , .n. 43 n . / 2 ‘és I D J I II nun-m [NIVMIIS mm WNW 61 Figure 4.9 Monthly Distribution of Labor Use Per Household in Small-Scale Fishing and Processing Production in Sierra Leone, May 1974-April l975 A. 96 95 39 Afi— "" n MARINE FISHING: NORTH l9! "3 J A S 0 I 0 J F I I B. INLAND FISHING J8 Mil-MM VWIDMINIS 62 Figure 4.10 Monthly Distribution of Labor Use for Small-Scale Industrial Firms in Rural Sierra Leone, May 1974-April 1975 A. CARPENTRY B. BLACKSMITHING 150 I“- u 1 120m 36 r: 3°22- fifi gm no '3‘— 550 #555‘55 fifi ., 3 so 29 JJASOIOJ"" uaaasonoarna ma mm C. TAILORING U ”LIB-1 70 g w .350 ”do 39 g. ,. 20 to 63 'l:;Z 1.l.--Peah Periods and Slack Periods for Selected Enterprise in Rural Sierra Leone 1|?~_ e 1 .1"? .. x-:-1 rt: :4 3:: an n.a.:luca. ww:tm1r~—J'u“j“_l h .. .19- Region Peak Month Peak Period -..J‘YY av..- Tasks YT—r.".-_£'fi'v'1—7‘flr“x‘f E'LF‘T Slack Period "3r:r”.c Rlvtrrdlfl (He;n.) Boliland (Mech.) 3*‘13"c fdardl Crosncvsts Cassa.a Onions-Peppers-Tonators tree 1(‘3 l rggs Cocoa Cof‘et Oil Pal" Oil Palm Egyghpn Fishing Marine Inland Small Industries Carpentry Blacksmithing TaiIOrinq Labor Sold Out "12m ZU‘MM Nat Nat Nat Nat Nat July Oct July May Aug Jan NOV Nov Dec June June May Feb Sept Jan Apr Dec Oct Apr July May May Nov July June-Nov. June-Oct June-July May. July-Oct Jan July-Aug. Dec Sept, Jan Jn. Oct-NOV July-Aug, NOV. May-Aug. Dec. Maj‘AUg Apr-Way. Seat Maj-5901 March-Aug Jan-Apr Aug-Oct Dec-Feb March-May May-Dec Sept-Nov March-Apr Nov-March. May-July Feb. May-July May. Oct Nov-Jan July-Aug, Nov-Jan. March Plant. Heed. Harvest Plant. Heed, Harvest Plant Land prep. plant. transpe. weed. harvest Plant. Transpe Harvest Transpl. harvest Plant. Birdscaring harvest Plant. harvest Land Prep, Plant. harvest Land prep. plant. weed. harvest Land prep. plant, harvest Plant. Heed. Harvest Plant. Heed. Harvest Plant. Heed. Harvest Underbrushing. harvest Underbrushing. harvest Harvesting. processing Harvest. processing, Tool Repair Dec-March Dec—April Sept-Oct. Nov-Feb. Apr-May June. Nov-Dec Feb-Aor. Oct. Jan-May March-Maw. Nov Feb-Apr July-Aug Jan-Apr Oct. Jan-March ”Ow-Mar:n HOV-Marcn. Jul. DeC 'ADT‘. Sept-Dec June-Oct Dec-July March-Oct July-Nov. July. NOV. Apr-July. Dec. May-Nov Aug. Oct. Apr Dec July-Sept. Feb ADr-July Feb. Apr-June 64 Inland swamp rice, shown in Figure 4.3, is characterized by sharper peak and slack labor periods than is upland rice. In the South and East, peak periods are in May (land preparation), July through October (planting, transplanting and weeding), and January (harvesting). The peak period in the North is for land preparation and planting in July and August; peak periods are more acute in the North than in other regions. The other four rice systems are characterized by two peaks-- one for planting and the other for harvesting. Fundi and groundnuts are generally planted one to two months before upland rice and har- vested one to two months before upland rice. Cassava,like fundi and groundnuts, is a subsistence crop. Cassava'a particular advan- tage is that it can be stored in the ground for long periods of time. Three farm enterprises--onions-peppers-tomatoes, coffee and wild oil palm--are undertaken in the dry season. The first two are cash enterprises while the latter is both a subsistance and a cash enterprise. The peak period of cocoa, also a cash crop, is during harvesting (August to October). Most of the nonfarm enterprises have peak periods between December and May. the period of low agricultural activity (Figures 4.9 and 4.lO). Labor use for inland fishing peaks in March and April. Tailoring labor is highest in May, the month of an important Moslem festival. Labor use in blacksmithing is heavy in May when tools are made and repaired for the planting season. Labor sold out 65 peaks in November and January for the South and East respectively. Other nonfarm enterprises are pursued most actively during the cropping season: marine fishing peaks in October, carpentry in July, and labor sold out in the North in July. Seasonal labor profiles for individual crops are important in explaining enterprise emphasis. In the East, for example, coffee and cocoa areimportant cash crops and upland rice and inland swamp rice are important subsistance crops. Coffee and cocoa have similar variable andcapital costs. Although coffee has returns per manhour about half those of cocoa, it is farmed by almost three times as many households and specialized in by almost twice as many. Coffee's dry- season peak labor period complements the rainy-season peaks of inland swamp rice and upland rice, whereas cocoa's peak period conflicts with those of the rice enterprises. High-decile farmers, who can better afford to divert scarce labor from food staple production dur- ing the peak season labor periods, devote seven times more labor to cocoa than low-decile farmers (Table 3.6). The ratio for coffee, on the other hand, is only l.8. Poor farmers, having fewer resources to fall back on in the case of staple-crop failure, are reluctant to pursue enterprises which conflict with production of rice, the food staple. Moreover, they lack the means to hire the peak season labor necessary to harvest cocoa. High-income farmers, on the other hand, have sufficient resources to farm cocoa, the most profitable tree crop. 66 The analysis of seasonal labor profiles is also useful in explaining the importance of the onions-peppers-tomatoes enterprise. In Chapter III, it was noted that onions-peppers-tomatoes is a low returns enterprise which is emphasized by middle-income households. Onions-peppers-tomatoes are grown primarily in the Freetown area, where a ready demand for vegetables exists. The two other principal enterprises in this area are upland rice and marine fishing (Spencer and Byerlee, l977). Onions-peppers-tomatoes' peak period is during the dry season, January to April, whereas peak periods for marine fishing and upland rice are during the rainy season. Thus, although onions-peppers-tomatoes is a low-return enterprise, it provides an important supplement to household income during periods when the opportunity cost of labor is relatively low. Furthermore, in the Freetown area, land for cultivation is scarce. Hence, another reason for onions-peppers-tomatoes' importance is its high returns to land, as shown in Table 3.4. Enterprise returns, as computed in Table 3.3, may not cor- rectly measure the profitability of an enterprise. If peak season labor is indeed a major constraining factor to increasing income, then the opportunity cost of labor varies between peak and slack seasons. Therefore, it is incorrect to value a manhour of peak season labor at the same value as a manhour of slack season labor. In Table 4.2 returns to peak season labor are compared with conven- tional enterprise returns to labor. The returns to peak season labor for a given enterprise are computed in the following manner: 67 TABLE 4.2.--Conventiona1 Returns to Labor and Returns to Peak Season Labor for Major Enterprises in Rural Sierra Leone Conventional Net Returns to - ~ . a Net Returns to Peak Season Enterprise Region Labor (cents Labor (cents per manhour) per manhour)b FARM Rice Upland Rice S 7.7 3.0 Upland Rice N 6.9 3.2 Upland Rice E 10.8 3.0 Inland Swamp E-S 15.8 24.0 Inland Swamp N 11.1 13.4 Mangrove S 15.8 24.0 Riverrain rice (mech.) S 23.8 30.6 Boliland rice (mech.) N 35.7 72.9 Other Annuals Fundi N 5.4 0.9 Groundnuts S 5.9 1.6 Groundnuts N 12.2 18.6 Cassava S 23.7 43.8 Onions-Peppers-Tomatoes N 10.0 n.a.b Tree Crops Cocoa E 33.5 51.9 Coffee E 16.8 n.a.b Oil Palm S 28.1 111.7 Oil Palm N 16.0 25.7 NONFARM Fishing Marine Nat. 36.8 44.7 Inland Nat. 64.7 66.5 Small Industries Nat. Carpentry Nat. 12.1 40.0 Blacksmithing Nat. 27.7 -- Tailoring Nat. 32.1 -- aRegions: s = South; N = North; E = b July to August; South: N.A.: Not available. For definition-see text: determined from Figure 4.1:Nationwide = June June to November; East: East; Nat. = Nationwide Peak seasons are subjectively to November; North: June to September. For these enterprises the values are infinite since no peak season labor is used. 68 Returns to Net returns to labor - (nonpeak season labor hours x peak season = wage rate) labor Peak season labor hours All-household peak seasons were delineated for each region. In the formula above, the higher the percentage of labor devoted to nonpeak labor hours, the greater are peak season returns. But this measure has several weaknesses. First, precise delineation of a peak period is somewhat arbitrary. Second, this formula assumes that peak season labor is the major constraint to increasing returns for all enter- prises. Third, the measure is meaningless for enterprises which are not pursued at all during peak seasons. And last, the assumption that all nonpeak labor hours can be valued at the average annual enter- prise wage is not tenable; this leads to an underestimation of the returns to peak season labor. In spite of these problems, however, the measure is useful wheninterpreted ordinally. The return to peak season labor serves to compliment the conventional enterprise return by taking into account seasonal labor constraints. Onions-peppers-tomatoes and coffee are low-and middle-return enterprises according to conven- tional returns analysis (Table 4.2). However, they have no peak labor requirements; therefore, their returns to peak labor are theoretically infinite. Other enterprises with high values include wild oil palm (South), mechanized boliland rice, and inland fishing. Upland rice, fundi, and groundnuts (South) have the lowest returns to peak labor since peak season labor requirements for these crops coincide with household peak-labor periods. 69 The relationship between seasonal labor inputs and enter- prise choice will be further examined in Section 4, which concerns enterprise combinations. The examples in this section have emphasized the importance of labor conflicts and complimentaries between rice, Sierra Leone's staple crop,and other enterprises. The peak periods for upland rice and inland swamp rice are during the rainy season. Fundi, groundnuts, tailoring, and labor sold out (in the North and East) have peak labor periods which conflict with the two major rice systems. Cassava, cocoa, marine fishing, blacksmithing, and carpen- try have somewhat overlapping peak periods with rice. Onions-peppers- tomatoes, coffee, wild oil palm, inland fishing, and labor sold out (in the South), have peak periods which complement those of the rice systems. Since the government promotes inland swamp rice at the expense of upland rice, it is useful to examine the degree of con- flict between peak seasons for each of these two crops and those of other enterprises. Table 4.3 shows that in the South, peak-period conflicts increase for three of four supplementary enterprises if a change from upland rice to inland swamp rice takes place. This is most striking in the case of cassava, which hasthe same peak month as inland swamp rice. In the East conflicts increase significantly with a changeover from upland rice to inland swamp rice because the peak months for coffee and cocoa conflict with the peak period of inland swamp rice. Conflicts also increase for two of five enter- prises examined in the North. Thus, in general, the seasonal labor 70 TABLE 4.3.--Months of Peak Season Labor Conflicts Between Rice and Selected Enterprises in Rural Sierra Leone Number of Months of Conflict South North East Upland Swamp Upland Swamp Upland Swamp Rice Rice Rice Rice Rice Rice Groundnuts 1* 2* 4* 2* -- -- Cassava 3 3*** -- -- -- _- Fundi -- -- 3* 2* -- -_ Onions-Peppers-Tomatoes -- -- I O 0 -- -- .Oil Palm 0 1* O 1* -- -- Coffee -- -- -- -- 0 1* Cocoa -- -- -_ -_ 0 3* Labor Sold Out 1* l 2* 3** -- -- NOTE: Computed from Table 4.1. *Peak month of one enterprise occurs during peak period of other enterprise. **Peak month of each enterprise occurs during peak period of other enterprise. ***Peak month for one enterprise coincides with peak month of other enterprise. 71 requirements for upland rice are more compatible with the existing farm system than are those of inland swamp rice. This conclusion must be accepted, however, with two reserva- tions. First, it might be argued that the household's selection of nonrice enterprises is subsidiary to the selection of rice enter- prises. If this is true, then conflicts between rice and nonrice enterprises are less important. However, since nonrice enterprises account for about 40 percent of total labor and 60 percent of income it is doubtful that their importance can be dismissed easily. Second, it might be argued that there is nothing immutable about the peak labor periods of individual enterprises and that changes in peak periods caused by the introduction of inland swamp rice may even be desirable. Although this is indeed possible, it is doubtful that the traditional agricultural system, based on upland rice production, would be characterized by sub-optimal timing patterns for the carrying out of important tasks. 4.3 Enterprise Variable Costs and Capital Costs The lack of capital to purchase technological inputs is a well-known constraint to increasing the productivity and incomes of poor rural households. This condition is one of the major reasons cited to support the introduction of credit facilities in rural areas (Tinnermeier, 1976). A related hypothesis is that the capital constraint prevents poor farmers from adopting high returns enterprises. Matlon, in his study of three villages in Northern Nigeria, found that the cost of 72 purchased inputs constrained low income farmers from producing some of the highest-returns crops. He also found a higher frequency of nonfarm occupations requiring high inputs of working capital among the higher-income groups (Matlon, 1977). In the analysis which follows, four cost components are presented to examine the capital constraint: annual capital costs, annual variable costs, the value of capital stock, and annual cash variable costs. These cost components, along with expected labor- 1and and capital-labor ratios, are used to analyze the effect of the capital constraint on enterprise choice. 4.3.1 Annual Capital and Variable Costs Annual capital costs and variable costs per acre for major crops in Sierra Leone are shown in Table 4.4. Annual capital costs are computed using the capital recovery formula (Liedholm and Chuta, 1976): _ rV R‘i-(i+r)-n where: R is the constant annual capital cost V is the original (undepreciated) market value of the asset r is the discount rate (10 percent) n is the expected life of the asset. This formula converts capital stock data into an annual capital cost flow reflecting both depreciation and the opportunity cost of capital. 73 TABLE 4.4.--Annual Variable Costs and Capital Costs per Acre for Major Farm Enterprises in Rural Sierra Leonea Costs Per Acre . Annual Farm ggglgble Cost of Total CapitalC Rice Upland 15.63 0.32 15.95 Inland Swamp 23.51 1.51 25.02 Mangrove 25.77 1.26 27.03 Boliland (hand) 9.10 0.50 9.60 Boliland (mech.) 17.00 0.50 17.50 Riverrain (mech.) 11.77 0.40 12.17 Other Annuals Fundi 7.29 0.28 7.57 Cassava 2.07 0.25 2.32 Groundnuts 10.96 0.38 11.34 Onions-Peppers-Tomatoes 44.27 1.19 45.46 Tree Crops Cocoa 2.38 9.53 11.91 Coffee 2.71 9.53 12.24 Wild Oil Pa1m - -- - aAverage figures for cases in which enterprise contributed more than 10 percent of total household income or labor. bIncludes costs of seed, fertilizer, land payments, hired labor, mechanical services and other inputs. cComputed using the capital recovery formula shown in text. For cocoa and coffee, the annual cost of capital includes an estab- lished cost factor which is comprised of a depreciation and an inter- est component. For other enterprises, the annual cost of capital refers to the annual costs associated with the use of tools and equipment. 74 Tools, equipment, and the establishment of tree cr0p orchards are the most important capital costs. Annual variable costs include both actual and imputed costs for seed, land payments, fertilizer, hired labor, and tractor rent. The crop with the highest variable costs per acre is onions- peppers-tomatoes, 44 Le. per acre, reflecting high rental payments and seed costs. Inland swamp rice and mangrove swamp rice also have high variable costs reflecting the extensive use of hired labor. Coffee, cocoa, cassava, and fundi have the lowest variable costs per acre, all under 10 Le. Annual capital costs per acre are highest for cocoa and coffee, 9.53 Le. per acre, because of high establishment costs. Annual capital costs are low for other crops, reflecting the rela- tively minor importance of tools and equipment in traditional agri- culture. In Table 4.5, capital costs are computed on a per-manhour basis, leading to somewhat different results. This method takes into account the differing levels of labor intensity of different enterprises. For example, although onions-peppers-tomatoes has the highest per acre costs, an acre of onions-peppers-tomatoes absorbs six times as much labor as does an acre of upland rice, and still generates higher net returns per manhour. The enterprise with the highest variable costs per manhour is marine fishing, 13.6 cents, followed by mechanized boliland rice and mangrove swamp rice. Annual capital costs per manhour are high for marine fishing, the small- scale industries, cocoa, and coffee. .N.e epoch to» mono: ammo 75 Al.1|1111 IT ... TI o.~ ~.P m.~ .. u- i- -i u- i- u- i- -- acmzuwsmxum—m ¢.v m.~ m.~ .. -- i- -- -- u- i- -i .. ocmuchmu m.e m.m m.. -- -- -- -- -- -- -- -- -- mcpcoppo» mmpcumscca ppmsm o o o -- -- -- -- -- -- i- -- .. ucmpcm m.- N.m o.mF -- -- -- m.- N.m o.mp -- -- -- oepco: mepgmpu zammm3 _._ o F.F -- -- -- F.P o _._ -- -- -- _ocau m_u:::< cacao o.¢ _.o m.e -- -- -- -- -- -- o.e P.o m.¢ A.zooev =_~Lco>_¢ o.m N.o w.m -- -- -- o.o ~.o m.w -- -- -- A.;ooev ucap._om p.m N.o ¢.N -- -- -- F.m ~.o m.N -- -- -- Aocasv oeap._om N.e _.o _.e -- -- -- -- -- .. ~.e P.o F.< m>ocmcax m.~ _.o m.~ o.m P.c m.~ m.~ F.o ¢.N o.m P.o o.~ asazm agape” o.~ _.o m.~ e.~ o a.~ N.N o ~.N o.m P.o m.~ u=a_a= mu; z¢ —m:cc< mpnmwcm> Poacc< opnmwcm> Paacc< opnmwgo> _c:owumz “mum nucoz zuaom .I.'.IU h i I u l W.“ N HIHIHIIUHHZIIN IIWHJI..“.I“"III""""H1|.F.HIW u ...H 11..” u . u n U n.1,.“ H III H.141H.'l.l|.“.u." "1"!“ All II II. magnescoz\mucmuv meow; mcLo_m Pocam cw mmmwcncmucm Lone: Low Czeccmz Log mumou puamamu ucm mumou opacpcm> pascc0 mum: mvpogmmao; o— :05” mmwp uczu muuuwvc0 mxcmpm .zmou :. so» u_oa on on wussmmm mew mmmwcacmucm Ecmuco: co» mu=0c_ ..o on ”muopa 000» 0.0ouaaouca on mucwsxca smou wEO0 gmzozupo mucwcoasou smog ocmu «>0: cu umsammu wgo mucwexma temp ppm--mucwsxma tam. Amy wzmou =_ co» v_oa 9L0 mmuw>cmm .muwcogume FFmiummu_>cmm puuwcmguws Acv "£000 :0 Lo; vwoa m? cmuwpvugaw p—o --Lmnp—mucm» Amv "uwmozucaa mew mmoumeouimcoaawaimco_co to» 06—00 ~F< .wuc—u wxmu cu .Lm>mzo; .mmmmgucaq uwmm meow .ucmcoasou zmmunam a w>mz cu umsammu mm: .mwoumeouimcmanma-mco_co mo cowuawuxw wcu :00: 00000 ummm __< Amy "Awno_v uumccoa soc» 00 0000 00:0 .mumou mpacwcm> 0o ucoconeou acouLan0 9002 on» so» an .coom_ umcwz mo ucwcoasou smog mzu oc_ccmocou umuum—pou we; came--conm— vmcwz Apv va=_ucp mumou m—nm_cm> mo mucmcoasou swoon 753 e._ .0 m.— m:050020xumpm 0.. 000 0.. 00000000» mmwcumzucw p—osm o o o o vcmpc~ ..0_ 00 0.00 00 0.0. o=_cuz 0=_00_0 z0<0zoz 0.0 00 0.0 0.0 00 0.. 0.0 00 0.0 0.0 00 0.0 2000 000 0.0 00 0.0 0.0 00 0.0 000000 0.0 00 0.0 0.0 00 0.0 00000 mmocu out» 0.0 00 0.. 0.0 00 0._ 00000200-0Loa0o0-000000 0.0 0 0.. _.0 0 0.0 ..0 0 0.0 0000000000 0.0 0 0.0 ..0 0 0.0 0000000 _.0 00 ..0 _.0 0 _._ 00000 mpoacc< cacao 0.0 00 0.0 0.0 00 0.0 A.00020 0000Lo>_0 0.0 00 0.0 0.0 00 0.0 n.00020 000.0000 0.0 00 0.0 0.0 0. 0.0 Aceagv 0=0___00 0.0 0_ ..0 0.0 0. ..0 000.000: 0.0 0, 0.0 0.0 0_ 0.0 0.0 0_ 0.0 0.0 00 0.0 05030 000.00 0.0 0_ 0.0 0.0 0_ 0.0 0.0 00 0.0 0.0 0_ 0.0 000000 0000 2000 0:0:ccz . czoccoz gnocccz . czocccz czcscoz czozccz 0:02:02 gaozccz cc; uwou 0000 \00000 cw; 00:0 sumo \xumcu L00 00:0 smew \00009 cog uwcu swam \mumou 0_00_00> 0000 . o_00_cc> 0_0c000> 0000 . 0_00_te> o_0e_.0> 0000 . 0.00,Le> 0_00020> £000 0 0_000200 0003000002 0000 00202 00000 Ilaiiu“F"lFH“IE.VNNhIPh[-wfihbh9.1.1:“...9 linnrlyuihnlll caoncez\0acou mcco0 cgcm_m 00:: :_ mompcugmucm Lena: Lo» unoccox cma mumou wpaowcm> smcuiu.~.v m4m $0 m_m»_m:0 $0 0000 0 0:00: F0>0F 0:00:00 _ 0:0 00 ucmuw$wcmwm 0:0 0:00000 :00300: 00:0:mmwwoke .00:0P:0> $0 0000_0:0 $0 0000 0 0:00: Fm>0p 0:00:00 m 0:0 00 “cmuwwmcmwm 0:0 0:0000: :003000 00:0:mewwok .xuopm quwamu $0 000—0> F0000 0:0 00 0:000: =P000000=0 .u:0F0>0=00 :00::0E u 0:20 .xpw>wuumammg 0.: 0:0 0.: 00_00h 50:» 00:_0000 00 00000: Lonmpipmpwamu 0:0 0:0F-:000_ mmwgagmucm .e.¢ :0wuumm :0 00:0000 00 00000: Lonmpipmumamu 0:0 0:0piconmp nmuuqumm mm.~ MN.F in in chowumz kmm. imm. nu ii umwm kmm.~ kmw.~ ii in sugoz «on. «mu. :1 in npzom mmmwcmgmp:m pp< ON. ON. Nvm N¢c pmcowpmz «rpm. krom. kkm¢m **omm “mam «kmp. «#mp. *«Fom ¥«NNN cugoz «rem. **PN. *kmmm **omm :paom xpco mmmwgacmucm 5:00 00000: Loampupmuwamu 00000: :000_-0000000 00000: 0:00-00000 00000: 0:0—-:000_ 0003:0000: 0:00: 0003:00002 0:00: 0003:0000: 0:00: 0003:00002 0:00: 0:000:000Caov :oto.\00=zv 00000: 00000-0000000 umuumgxm 0000: 0:00-:0000 umpuquu 00:00; 0::mwm F000: :0 :00000 00 00000: 00000-0000000 0:0 0:00-00:00 000000xm-..0.e mgmééé§éé 4 g,/////// W// ///// //fl 2 700 5" 200 96 C. luscious ( [Scam ) Inland tiu- fundi "III E mm mt IIIEI mmlms - n u'.'.l‘.-" .‘ .‘..-' -~ -: 400 3,." ' 0 ,v . '5"- In“: I. lonMsUIcw) mm rice-tailoring. Cam-try HZ . . .v . ...:"Qtlu ‘ .\ I‘ rifle-9.2: ‘4‘...'...’f,..: _- 01‘.,'."' ‘ ' .n .- .‘J “5 . ...”: . . D '. ‘ 2” 25 17 1 ':."\ o"- _s-. 2&1. . 16 [III/II zzzzzazayzq I M 25 O loath ‘3'?! .'9‘ Q2. 25:534. .:-. 5v. ‘1’?!“ q IIIlIIIIGJIOIIOII Q "till I!“ 2113mm as m El /» Q3 ,zll,h552222 f V/I/I/ , 97 I. mum ( Hum ) um tiu- mm mm: C] mm m: OIIEI mums“ m m s» m ‘“ ’ m m 4 m m -..- ': -'r . .\}”"‘”}:.u ”"3. i"; {'21 35,-3.1. ° '; r»;- '- avg-{p.232 - .° ' . u, ' on ' 0 .'~ t m b: 331 m m w ‘ V : ‘- L . 3 - ‘A 3"." o . 5".-"-~.-‘,’.-.‘ '(D. '3’. o. 2» 5323.323. j...“ 1.“. .l. l" 7 1' 1" r /,// // ////// '/1§§5§7 .é './/// //. 1299" l l I S 0 I I I F I I >oz o.~m xmz-umo .>_:n-mc:a mean Pouch upocmmzo: .9 _.mm xmz-umo zpsqnmcaq mczc m>mmmmuumuwa acmpa: .u mag .soz-amu owe.»P=a ~.oe >oznpnmm xmz-ga< zmz m>mmmmo .m .>oz-auo me «.me xmz-umo xpacnmcac mesa muwm ucmpga .< m>ommmu-muwm wanna: .¢ m.- La<.nmm-umo >oz-m==a poo-»F=q Pouch u_o;mm=o= .o m.ne ga<.ammnuwo uuo .zpaw uuo muscucaoswumuwm campus .u m._m gugmz->oz ynmm-a=< apan m:=w-ga< uqmm musccczogw .m m.m¢ sz-gaoz-apsa m=< Page? aposmmaoz .o >oz-puo o._e xmz-umo m:<-»_=a xpan use upom gonmaumuwm ucmpaz .u m.m¢ mczaugn< suz.:ma->oz >oz use n_om Loam; .m m.m¢ xmz-umo >oz-goo m:<-m==c xpac mowm unwraz .< me “so upom gonm4-muwm ucmpn: .N ¢.¢N zmz-uoo >ozum==w zpzw page» upogmmzo: .o >ozuuuo «.mm xmz-omo m=<-m==w >_:e spam Fwo upmzumuwm vampaa .u m.me >oz-m==a amz-gg< ga< spam Fwo upwz .m >ozuuuo ,. N.om Amz-umo ms we vowgma xumFm uo_gm¢ xmma cucoz xmma acmP0w$wmoo meow; mggmmm ngsm :? mcomumcwnsoo wmwgagmucm gown: Low mm: Loam; $0 mvowgma xmwm new xumpmun.F.m mgm<~ 99 o.m~ xezueeo apee xFee Feueh upecmmeez .o m.~v xezueeo ueem-meee xpee .egeu..Fweh.mumm .e: .u m.m¢ Le<.em.ueo-eee gez.=ee.>ez seem: xguemegeu .mcmgepwe» .m F.e¢ sz-eme .eamm-m==e apae move ego—Q: .g xguemegeu OF -mcwgopwmk-m0wm agape: .m m.o¢ Leez m=oz e=<-m==e xpze Pecze-mqu agape: .e e.mp_ Le<-eemm e=<.m==e e=< _e==u .m e.me ga<->oz zpze »_=e «gem agape: .< mF wecemnumewm campus .w ~.mp zezuemo >ez .ecee >ez Pouch epegmmeez .o >ezuueo N.¢m aezuemo meezupeo m.mm Leezucegez ceeuueo can memmeo .m P.mm Le<.em.._.emo apee xpee mewm eeepen .< meemeo-eewm campus .e N.om xez-eeo m=ez ueemuxpee peem eewm eEeZm veepca .m m.~m >.e_,_.e<.een_.een_ >ez.»peuee maze eewm agape: .< em meme namzm ecmpce-muwm e=m_ee .m memeo :eweevge> we eewgee xeepm eewgee xeme gaze: game ecmmueeemoe .eeeewpceunu.p.m m4m€;T E""° Rfééj/{éjhgggq‘ 314‘;th $133, r2127: Z//////fi m Veer-<2»: 42° 4 14 " '/3y/jg.;4//.//, ”,7 383 3" 314 "”7 33s 3“ //52/ III III, 3.5 2,2 ,yflv‘i 2” m / 1‘35” 73,; 53(7) ////'5'- ggfl 3’?" :21", a??? m ““234 ‘21» III I I I I S 0 I I I I I I Ieeths I I 3 E. Ipleed riee- leleed seen. me Ieeeebelds (“as“) no [:1 felilyleber 7 1H ‘5, Iiredleber III 7477;: I" no 7 77 3:2 Iee .3237? 077 7:7 33’ / //’1'/ 7)”/A’// "4 e 777 554 7, 7 5.1 ..f 5.. %/,%;f/,/: III ’l/féf/ /,/77 / III “'5' "7'77 m r 477/ .7 . 5 7717/7; 7/ 7 4n 3" 432 /.% :34}, 3 e1 3 ‘ 2 . 77.7»; 3 37 / .17 3 e e 5""7L'L; 7296’ ” ’,’mn ugmccoum u co_uo_go> mo acm_uwmwoouo mesa xpaa o.a~ No.mm o om ~.m_ xgucwagau --mcwgoppoh--mU+m ucmpn: mm: »_=a m.m¢ gm.o¢ N m m.,_ Pecan--wu*x u=~.a= .>oz .m=< ~.op No.om 0 mm o.mp m=_;mvu--ou?m ucopga mcaa »_=n o.om wo.qm 9 ms m.e_ mm$wou--mu*¢ u=a_g= mesa »_=q ~.om N0.Fm op _~ “.9. mu_m asczm u=~_=m--mu_¢ new—a: .>oz x_=a o.~m Ro.m~ F Rm o.~_ ~>ommmu--uu.¢ u=m_a= .>oz .oaq m.h~ “a.mm N om P.vp muscucaogw--ou_m u=~_g= .>oz .o=< ~.- RQ.NN -- ~¢ “.mp uao u_om Lcamd--au.¢ u=o_g= .>oz .o=< e.q~ fim.em m mm m.e_ e_~g .wo u_?3--mo.¢ ego—a: Lona; goes; cmgw: agoncg «Loam; nmgw: Pouch pouch mo Loan; Loan; Loan; Page» »_mswu yo mmmucmugoa mmmucmugma mm —muoh mo cowumcwaeou mmwgggmucm umng Low >__Eou Low “comumwgc> “cowumwgc> we gonna umgmz Lone; vmgmz mmwucmugwa mm zucoz xoma sacs: sown mo ucmwuwmmmou yo acm_uwbwwou wmmgagmucm-ccm muwm uccpa: Loam; awn“: _ ”I. ..l I‘l‘n'l'Ul‘IIII to LO V M N y"!.'§h”fllfll.fi"h "HUI“..NuN blpnhfllfl H i‘nflluth “I.“ "MVP ..zsIN bu.“ .1th. u. ruuhlln u m ,clluu‘“ uh “kin!“ .llhll,’ . In H I ,N‘ H. I. acomA mLwam _mgzm c_ mcomumc_nsou mm_Laqucu Lawn: cw mm: Lona; uwng--.~.m m4m' v’r' ’- V . . A , .14 ’1 4y. ' "/.4 , ,. , ,' / 4 '. r,/.f,/’f/ ‘2, ‘3‘ 1 AIM/xx, :V/l’z :4 ‘11 r'. Jig/'3’ ' ' . 9" r/',‘ :4’?/’ ...I’ 'I'.’ r .,,A ;/.'l.,/ . /,/,/ O I _’ {I 5" /’ ’A ,/ 5 31 e s 1 249/4? ”/51 '~ 436 436 339/3? ’ , x/x /. //.-.»;’>3r , s. if fizz/Xx} ‘ 2' ,3 , , 0 'r .j/ _ e» N N \\ \ \ \ \ u - K N U N N l I I S 0 I D I F I I c. [on moon: lousmus [Z] lived later D fully later I” / /'/ . 4 ,. II. leleed rice-Oil pal-(end) 120 I. me IICIIE msmus e e e I 5" “3 m lindleber 4 4 3 4 31 421 / - : V :2: A , ‘ 1‘ ,r‘ 1 W //; [:1 rum ...... 4 u 353 ://////332;///37'521 ,, W gig/M; '7, H1141 //,1/,;’_1 3 e e “/2513"? 3‘ I m 13 304 3.. 162/ 261/ 2“. /Ar/ / //// 7 z“ / 2 57 1‘ 3 ./ 4. d7 zen 221 1 e e I I I I S 0 I I I f I I I. emu uncut lousmus 12 7 7 . 4 /' 77 1 . . 721; -. ‘ ‘ ‘ 3/52’6654/(1’33131 lired later 631 r/n “4'5 7; 6 25 A”3ié""“f/' T/ {/7 '1’. 5.x",,gi/,T/// “‘4'“ ,‘c' r '. ‘IZ %% I: . ‘ . . f/x'l/‘III’Z/ {1" i ‘2. ‘ u o y/ 'I.”’ ....f 0293/2? ’ 47% S 2 ‘ 552 s 3 , s e I 4 16 1M”; 433 W ”:2 /// . V/BJQE 52%;;A 4 e s r V”; 415, I ' 4H m 711;; 364 412 V316"; (/,'/,,//',/‘ 3 . . //'1’74‘/ 2 3 3 2 e e u o I I I I S 0 I I I f I I we 4 613 no 2,751.7, 57. 573 c. [OI mm: neurons 52} K14531114?!” ,_ Iired labor 544 .0413 522 ~24“ Z/V/f/xj/ 3// / 7 E ‘ . . @2474 {[47:22 fully labor 3‘! m 415 3 n 304 % m 215 "”3"” m 233 V, I, 24! fl / 3oz. ‘ AAA 1% / 1, 243 / 7 m 254 1 I I z 1 1 2 3 I 1 e e III. leleed rice-6333333 lZl 3 3 3 I. 31311 13m: 3335333133 53 0‘ 1.” 31:34 1.3.: wj/ ‘ Z”/ V 433 [:1 mm 1.331 7:5" lv'/./' Ar]; 7, , , ,7 ' '.,/ ”3‘34” 3" 4415:4332. 3“ /”// 233 3 3 3 24:32?"."’»'efl;2;4%4» 3" 4413"; 3“ 4‘4 1 ”2*;34‘ 232 II 23! 131 1,3 ". l {5’ 117771 I.” ///;-7., ”’7 113 1“ 1" 133 ' ”2212.4" 155 151 1 3 31 I I I I S 0 I I I f I I I. emu 11133.: 3333333133 333 74% mm mm 533 5471” , . . 7939 W}; [:1 mm, 1.3.. W/ , I82 ,//, 22 444911? ,3, 4m, -.,:///// 0 E442; “I:~////‘ /. r , II, '1}: 1 5": '/ "////:/4 ‘ 3 4. .1 313 /// 41 ,. ,. 449% 397 471;" “‘3 9;; a 3’3 "’ 39‘ 333 52:23:59“ 3'”. //%f 3" fl/ a 333 33 /////fi;/ // 321 3 3 3 3 33 ¢///4//////%/4@ %I%;Z 44,).///,,l,/,//// ,.,,.. l i A 233 233 1’1 2’ 233 2 3 3i 1 3 31 I I I I S 0 I I I f I I 3.1" 11133313 331133113135 .W 311341334: 3 3 3- ,5” 523 A}: 523 [:1 min, 1334: 506 réz/I . ,7; 1 5" 7777 _ {If 53 o ,, /// 3‘ 2:4 1.4573/ 1213!,» p’/1,,.,4 "—1 44: 07/. 4 g 5 9.137) 4A 717/11,» .-2 ’W $2775; 411 437 433 924/4 , 2'7 {175/ r//’/, /Z I 0 m r, [/11 r /1%/// 53 397 33s 5‘, j, 335 3 3 3'9””; / 233 £117 W/"? 1 311 . 7, 214 (£27,? W4 ’21!) 233:4 233 233 v 243 233 241 1 3 3 122 .mgmgonmp umgw; ucm mgwnswe xpwsmm Pym $0 Lonm— op mswmmg gonmp pmpop .mgmnsws xpwsmw ppm mo Loamp op mgmwmg Loamp zpwsmm usa:_ Loam, xpgucos came .oo_ x copgmp>mv ugmucmpm u cowummgm> mo ucmwuw$mwoum .>oz .m:< o.w~ ~.w~ o.¢F fiom amaze; .>oz xpaw m.mp m.mp ~.~N &o¢ mpuuwz 2.2, :2. m.mm 9% of “an :23: m>mmmmuuumowm ucmpaz .puo .>oz o.¢m m.mm N.o_ gem ummzoA .uamm >F=w F.~N o.m~ ~.op xoe m—nuwz .>oz .5 2m #8 mi Rom :23: mascucaoguuumumm ucmpa: mcaw zpza w.Nm ~.mm ¢.ep Nom pmmzoA .>oz .uuo N.mm m.e~ m.NF &o¢ wpcuwz .uuo .m=< o.~N N.om N.¢_ mom pmmsmwz anPqujspma F_o--muwm cannmm LoamA Pouch gonm4 Lona; agonmg Page» agonmA xpmsmm mo mam umng xpwsmm ”cowumwgm> yo ”cowamw;m> we nucmugma mm gzogw msoucH cage: xmma "gaze: xmma pcmwumwymou pcwwuw$mwou Loam; umng meow; msgmwm Pagam cw mcowpmcwasou mmmgagwpcm umpumme Low azogw weoucfi »5 mm: Lonmg umngun.m.m m4m