Zambezia (1998), XXV (ii).DEALING WITH IMPOVERISHMENT: SOURCING ANDSELLING FOOD IN MASVINGOVICTOR N. MUZVIDZIWASociology Department, University of ZimbabweAbstractSourcing and selling food are the main concerns of this article. Food vendingwas not a strategy to climb out of poverty: it was at best a coping strategy. Thefood-vending market niche demonstrated the existence of interconnectionsbetween food vendors and the formal and informal markets. The state andmunicipal authorities' responses to food vendors reflected narrowly-conceived,male-dominated, official development policies that marginalised the strategiesfor the alleviation of poverty by women in the informal sector. The wayofficials treated food vendors demonstrated the existence of competing andconflicting rationalities between male decision-makers and poor women.Those in authority in Masvingo seemed to be propelled in their actions by thedesire to control. On the other hand, my Masvingo respondents strove toresist and circumvent the system of controls put in place by the powerful. Myrespondents' concern was to engage in those activities that contributed to thelivelihood of their households.THE FOCUS OF THIS article is on sourcing and selling food by Masvingo'sfemale heads of households. This study is based on urban anthropologicalfieldwork over a period of 14 months from early November 1994 to the endof December 1995. Information based on fieldwork conversations,observations and interviews form the basis of this article. The Masvingosample was selected using snowball sampling techniques (see Muzvidziwa,1997). The results presented in this article comprise part of a largerproject.The study is situated in Masvingo, a provincial capital of some 52 000people (CSO, 1993, 13). According to the 1992 Census the town had anofficial unemployment rate of 25%. The majority of Masvingo's residentswere self-employed. For 38% of my respondents, food vending constitutedtheir main source of income. Another 18% engaged in food vending as asecondary source of income.My respondents considered sourcing food, for self-consumption andin the case of food vendors for selling, to be of prime importance in theirlives. One woman in the study noted 'tongosevenzera sadza chete' (wework for sadza Š staple stiff porridge made from maize-meal Š andnothing else).Tactical shopping arrangements and gardening ensured a steady supplyof food. Half the households produced for their own consumption, but147148 SOURCING AND SELLING FOOD IN MASVINGOTable 1MASVINGO RESPONDENTS' MAIN SOURCE OF INCOMEMain source of incomeFood vendingCross-border tradeFormal jobProstitutionCarrier-bagsHairdressersOther*TotalFrequency19136432350Per cent3826128646100"The main source of income for these three respondents were respectively maintenance,rentals and traditional beer brewing.only 4% of my respondents also produced for sale. Food procurementstrategies enabled these women to survive and cope in the city. To use theMasvingo (and Zimbabwean) colloquialism, many people ate 'air pies' (aeuphemism for going without food) for lunch.Hansen's (1980, 212) assertion that in Lusaka, Zambia, food vendingonly provided women with cash on a short-term basis and that it did notoffer long-term economic support or benefits, was not applicable to thefood vendors studied in Masvingo. Pooling of incomes was rare in mysample. Due to the limited number of adults in female-headed households,they found it difficult to pool incomes. Urban households, such as those ofMasvingo research participants, were at high risk, because unlike otherlarger households they could not put more members into the labourmarket in order to compensate for the general decline in incomes. In theabsence of income pooling and contributions to the household budget bygrown-up children, the women in the sample ended up overworkingthemselves. On average they worked more than 13 hours a day. In the longterm this could have a debilitating effect on their health, even if they hadeaten well.SOURCES OF FOODUrban sources of foodProduction of food was fairly widespread amongst the Masvingo researchparticipants. At least 50% of the participants produced food itemsconsumed by their own households. This was a very high proportiongiven the fact that only 14% were home owners (66% were lodgers and20% were tenants). Production of food took place in the house garden.V. N. MUZVIDZIWA 149The size of the urban garden was quite small, ranging from a singlebed approximately 6m2 to nearly 100m2. The portion of the yard put toagricultural use depended on the number of lodgers resident. A gardenplot was not a normal entitlement for a lodger. Its use was negotiatedseparately from the room space, and depended on the generosity of thelandlord. However, at least 36% of the respondents, excluding houseowners, had access to a garden plot. No extra money was paid by a lodgerfor the use of a garden plot apart from room rental. The water for gardeningcame from the house taps. A single water bill was shared amongst theoccupants irrespective of whether they all used water for gardening ornot.The respondents who practised urban gardening grew vegetables,most commonly rape and covo. Only 16% of the women cultivated tomatoes.The women felt that tomatoes were difficult to grow, susceptible to manydiseases and required many inputs. Thus it was uneconomic to cultivatetomatoes on very small acreages such as the ones they had. During thewet season, in addition to cultivating vegetables, 40% of the respondentsalso planted maize in their gardens and two respondents used fairly largetracts of municipal land illegally. They utilised the vacant land oppositeOld Mucheke area (the authorities appeared to turn a blind eye to thisactivity). Although the cultivated land, which depended on natural rainfallunlike the household gardens, was nearly a kilometre away, no cases oftheft of crops were reported.Table 2 shows who cultivated these crops. Sometimes a boyfriendgave a hand, but labour was rarely hired for this type of work.Table 2SOURCE OF LABOUR FOR THE GARDEN PLOT (INCLUDING HOMEOWNERS)Number %26141050Total 50 100Despite the small-scale nature of these urban agricultural operations,of Masvingo's female heads considered their urban agriculturalactivities as an important aspect of what they were doing in order tosurvive in the city. Urban gardening provided the women with theirSelfChildrenFriends and Co-residentsNo gardening n/a137525150 SOURCING AND SELLING FOOD IN MASVINGOvegetable requirements. This allowed the women to budget their cash foritems other than vegetables. For the female heads, gardening played animportant part in balancing the domestic budget. Some women dependedon the house garden for nearly all their vegetable requirements, whichtranslated into meaningful savings. Money spent on vegetables by thosewithout gardens varied between $50 and $150 per month. It was also quitecommon for respondents to give vegetables to a friend or neighbour as agift. One woman noted 'It's worthwhile and you feel good when you giveeven two leaves (of vegetables) to an acquaintance'. Through gardeningthe women in the sample were able to expand and strengthen their socialties.Illegal cultivation was practically absent in Masvingo and confined tocultivation of unused vacant Council land. Only 4% of respondentscultivated illegally and only in Mucheke. They paid no rent to grow maizeon Council land during the wet season. During the 1994/1995 agriculturalseason, one of the two illegal cultivators in my sample managed to reap six90-kg bags of maize. This was likely to meet her staple food requirementsfor her urban household of eight persons for at least a year. The secondreaped three bags of maize, or a four-to-six month supply. They groundthe maize at hammer-mills that have sprouted in Masvingo's residentialareas.Rural sources of foodOnly 8% in my sample had access to rural land, yet 20% of respondentsassisted kin and children with agricultural materials, especially fertilisersand maize seed. None in the study sample themselves produced fooddespite the fact that 72% of the sample maintained roots in both town andcountry. There was, nevertheless, a small flow of food products from thevillage to urban residents. Whether respondents were double-rooted ornot, at one point or another most sourced food items from rural petty foodproducers and suppliers for some of their urban households' foodrequirements. The female heads in the sample noted the sources of theirhousehold food for the month preceding the first interview. Seventy percent indicated that they had bought some food items directly from ruralproducers or suppliers, such as green maize, groundnuts, roundnuts,beans, sweet reeds, mushrooms, home-grown fruits, wild fruits, vegetables,sweet potatoes, peanut butter and unground maize. Some of the womensourced these food items directly from their villages, during trips home.Donations and gifts of foodFood donations and gifts came from kin in the rural village and in town.Many women in the study received food from village kin when they visitedthe village, and when kin visited town they brought food items as gifts andV. N. MUZV1DZIWA 151for use during overnight or short-term stays. Within the 12 monthspreceding the interviews, 34% of women in my sample had received fooddonations from their rural kin.The flow of urban food donations and gifts, though small, was mostappreciated and came from a wider cross-section of people. Everybody,with the exception of two female heads of households, received fooddonations from urban friends. Another 30% and 28% received fooddonations from urban kin and neighbours respectively. Generally fooddonations were irregular but came in times of need reflecting the generalpatterns of how the women coped with urban existence.During times of need, religious organisations also made food donationsto some needy members. Religious participation was quite high for theMasvingo study population and 62% claimed to be active church members.Masvingo churches assisted some of their members with food donations.However, this required a public declaration of one's incapacity to feedone's household. The women regarded such public declarations asembarrassing, which many of them did not wish to undergo in order to beconsidered for food donations from the church. Other than at funerals intown, assistance from the church was thus very low. During the yearbefore the initial interview, only 14% of the study population receivedfood donations from a church.Food provisioning sources: Purchasing food in townAs shown in Table 3, most of my respondents purchased their food itemsfrom both the township and town centre, and also from formal and informalmarket operators, with a very heavy reliance on informal food suppliers,which was an effect of the liberalisation of food procurement policies dueto the Economic Structural Adjustment Programme (ESAP).Table 3SOURCES OF FOOD PURCHASES INFood vendors, groceries and supermarketsFood vendors and supermarketsFood vendors and groceriesTotalTOWNTotal358750%701614100Food vendors were the most important source of food supply for mostwomen in my study. Research participants who did not sell food themselvesnoted that they normally purchased from a particular vendor or group of152 SOURCING AND SELLING FOOD IN MASVINGOvendors. They had cultivated personal relationships that transcendedeconomic transactions. Thus over time, for most women, the market andlocal grocers ceased to be impersonal. Buyers and sellers became sociallyfamiliar, exchanging information, greetings and sympathies. On the otherhand, town supermarkets were impersonal. Only the food vendors in mysample bought food (generally for resale) from bulk-food markets, wholesalemarkets and mobile lorry suppliers.HOUSEHOLD BUDGETING TECHNIQUESFood provisioning was linked to varying levels of budgetary skills by thewomen in my sample. Budgetary skills were a key factor in successfulcoping strategies especially in the face of scarcity. The management ofsmall erratic incomes was a daunting task that affected food consumptionpatterns. Very often if not all the time, most women in the sample had tostretch their limited food supplies to last longer than actual supplieswould normally allow. This required special skills. Reducing the numberof meals per day, occasionally skipping some meals, reducing foodquantities served at each meal, were strategies used to stretch foodsupplies. Stringent food budgeting allowed my respondents to survive.Most indicated that they were simply limiting purchases to basic foodrequirements.Despite the fact that 96% of the respondents kept no records ofhousehold expenditure, through their experience of a hand-to-moutheconomy, they exhibited a high level of financial and budgetarymanagement skills. One divorcee and a mother of three noted, in connectionwith keeping records of expenditure,I don't want to be reminded of my sorrows and everyday failures. Tobudget and keep records simply adds to my misery. I don't want that, soI don't really budget. But it does not follow that I don't know my expenses.Another never-married single-mother pointed out that budgeting wasa frustrating experience. But according to her:It does not follow that if one does not budget, one is not sure of expensesincurred to run one's household. Despite the lack of a budget, fromexperience and by checking and keeping an eye on items in the home, Iknow what I need and how much I spend on a monthly basis.Eighty-eight per cent of research participants controlled their ownhousehold budgets, whilst 12% exercised joint budgetary control. Inaddition to the three respondents noted earlier who shared meals, tworespondents in mapoto relationships shared budgetary responsibilitieswith their partners. Another respondent who shared a room with her twoyounger brothers, also jointly controlled the household budget with oneof her brothers. There were two cases where children were responsibleV. N. MUZVIDZIWA153for controlling household expenses but my respondents still remained themain actors as supervisors and main income providers to the household.Joint budgetary control occurred in the few instances where there waspooling of income and in instances where children contributed to thehousehold food budget. In my sample, 86% claimed to know theapproximate amount spent monthly on food. However, I should point outthat I did not have the means to check the accuracy of such estimates andclaims.Food expenditure patternsInitially my respondents did not distinguish non-food items from fooditems in their budget, until asked questions that required disaggregatinghousehold expenses. All household items were referred to as mariyemumba('the money of the house' or household expenses). Housing-relatedexpenses (water, electricity and rates were generally shared expenseswith other households) were not treated as part of mart yemumba.However, no relationship was found to exist between household size,consumption requirements and the proportion of income spent on food.Table 4 shows that the proportion of income spent on the food budget byrespondents was positively related to the poverty index.Table 4PROPORTION OF INCOME SPENT ON THE FOOD-RELATED BUDGETAND THE POVERTY INDEXFoodexpenditureLess than 50%50% to 75%75% and AboveTotalChi-squarePearsonBurntoutO CM CM4Value24.58812Hangingon261220Coping Climbing5 710 22 017 9Significance0.00041Total14201650%284032100As expected, the more successful respondents spent proportionatelyless on food compared to the poorest. Many respondents had little or nomoney to spend on other goods and services. There has been pricedecontrol in formal sector produced mealie meal, creating an opportunityof all sorts of entrepreneurs like respondent S who sold hammermilledmealie meal.154 SOURCING AND SELLING FOOD IN MASVINGOResearch participants changed their shopping strategies to enablethem to cope with declining incomes. The majority purchased overripetomatoes in small quantities, to use as relish. Buying overripe tomatoes ata discount was a rational strategy, for only 10% had refrigerators, not all ofwhich were working. This meant that one had to shop carefully, buyingonly items immediately for the pot. Nearly all my respondents bought insmall quantities, mostly when the need arose. Only 8% bulk-purchased ona fortnightly or monthly basis. Hence most of the respondents paid morefor goods because they could not afford bulk purchases. Bulk purchaseswere made at supermarkets, but many respondents bought only singlediscounted items at supermarkets in town. In addition to priceconsiderations, purchasing from particular outlets was also influenced byconvenience. Those who worked in town did part of their purchases attown supermarkets. Respondents who shopped in town pointed out thatthey frequently compared prices and regularly changed from onesupermarket to another depending on price.Sourcing food invariably involved different shopping strategies,considered useful by my respondents as they struggled to adjust and copein town. For Masvingo female heads shopping was almost a daily routine,and an important aspect of their lives. As Kemper (1981, 222) observed,going to the market or to the store involves getting someone to watchone's house and one's children, to accompany the shopper to carry homethe purchases, and (upon returning home) to exchange some items withthe neighbours for those earlier borrowed. However, Masvingo researchparticipants went shopping mostly unaccompanied. Most female heads ofhouseholds did their own shopping and only a quarter got assistance fromtheir own children, maids or co-resident relations.Household organisation of meals'Organisation of meals' involved examining cooking styles, the number,frequency and quality of meals, and waste and conservation strategies.Most women in the study spent much time, effort and energy in organisingmeals for themselves and their households. On a day-to-day basis,respondents in the 'burnt-out' and 'hanging-on' categories had to stretchtheir meagre resources in order to ensure there was some food on thetable for themselves and their dependents. The tendency of poor peopleto eat less during meals and eventually to reduce the number of meals perday, reported in other parts of the world, was also found in Masvingo.These were deliberate attempts to reduce costs among my poorestrespondents.The use of the word 'meal' can be misleading. Most of my respondentsdid not eat real 'meals'. Instead the Masvingo women defined the eating offood items at certain times as a 'meal'. Thus a bottle of coke and a slice ofV. N. MUZVIDZIWA 155bread around noon is lunch. Often this type of meal was all that constituted'lunch'. The question the women were asked was 'Munodya kanganipazuva"! (How many times do you eat a day?) All seven who ate threemeals a day had incomes above $ 1 000 per month, and were categorisedas coping with (three) or climbing out of poverty (four), and of these fourwere cross-border traders.NUMBER OF MEALSOne meal per dayTwo meals per dayThree meals per dayTotalTable 5PER DAYBY RESPONDENTSTotal538750%107614100Food procurement and cooking practices were important in loweringthe cost of living for low-income households. In Masvingo cooking practicesensured minimum waste. From my discussions with the respondents I wasstruck by the absence of cold meals. Left-overs were either integratedwith new dishes or were warmed before consumption. The issue of warmingleft-overs could be related to what has been termed the tragedy of thecommons with reference to communal rural resources. It is important tonote that the energy bill just like water was mostly jointly shared. It wasquite clear that where people shared costs there was no motivation toreduce expenditure, as was the case with my respondents. Mostrespondents ate warm meals, so meal preparation had to coincide withthe time when most members were present, in the mornings and evenings.Generally the meals were simple, not too demanding in terms of theingredients used and time required to come up with a complete meal. Themost commonly used food items were multi-purpose tomatoes. Tomatoesfunctioned as part of nearly every relish, or as relish on their own.Many households ate poverty diets leaving a lot to be desired in termsof nutritional status. Table 6 shows that 22% of respondents ate onlysadza (thick maize porridge) and vegetables most of the time. Fruits wereabsent, making the diet for 66% of respondents inadequate. From thesample it appeared only 34% of households had a varied diet that consistedmostly of sadza, vegetables, meat, occasionally rice, sometimes fruit, fish,and beans.117151722143034156 SOURCING AND SELLING FOOD IN MASVINGOTable 6DIETARY TYPES FOR THE RESPONDENTSTotal %Sadza and VegetablesSadza, Vegetables, Tea and BreadSadza, Vegetables, Tea, Bread and Meat*Sadza, Vegetables, Beans, Fish, Meat, FruitTotal 50 100*Also had additional food items from time to timeThe frequency of meat intake was taken as an indicator of the nutritionalstatus of the household in the sample, given the fact that Zimbabweansgenerally are meat-eaters. The research participants were asked to indicatehow many times they had taken meat during the course of the weekpreceding the initial interviews. During that week, 22% had eaten no meatat all, although none of the women in the sample was a vegetarian, while28% had meat at least once but not more than four times. Absence of meatin the diet was a result of problems of affordability, although half claimedto have eaten meat five times or more. Almost without exception, thewomen ate 'ration meat', shin and tripe, low-grade meats. A commonstrategy amongst the women was to stretch the meat supply by eatingonly a small piece at a time to give flavour to the vegetable dish. Based onthe few observations I made for a small number of meals served during mypresence, two kilograms of meat would have been adequate to serve fivemeals for a four to six person household.Vegetables were the most commonly used type of relish. At times forsome women relish consisted of tomatoes only. Except during the rainyseason (from November to March), there was very little variety in the typeof vegetables consumed by some respondents. They rarely ate rice orpotatoes. Table 7 below shows intake of these other items by respondents.It appeared that the general dietary pattern reported by respondents andwhat they had eaten in the previous week did not quite tally. This could bedue to attempts by some respondents to present a favourable picture oftheir dietary habits or simply a problem of counting.Waste and food conservation strategies adopted by Masvingorespondents reflected the cooking styles observed by Logan (1981,241) inMexico. A tendency to minimise waste and recycle left-over food was quitecommon in my research population. Some 64% of respondents recycledleft-over food whenever possible, but left-overs were not a daily occurrence.Twenty-six per cent of the respondents noted that they had no left-overs561415481012283096V. N. MUZVIDZIWA 157Table 7INTAKE OF FOOD ITEMS EXCLUDING MEAT IN THE WEEK PRECEDINGTHE INTERVIEWTotal %BeansRice/PotatoesFishFruitBread (mostly white)to recycle or to throw away, as they became more economical and efficientin preparing just enough for household consumption. There was a generaltendency for households to prepare just enough food for consumption atany one given meal. The female heads joked about how people went togreat pains to avoid wastage. The insane and dogs that used to feed fromwaste-bins could no longer do so: in Masvingo they rarely contained foodduring the period of my research.Left-over relish portions were generally integrated into the next meal,because most people did not have refrigerators. Left-over sadza wasrecycled as maheu (a type of traditional drink which, depending on degreeof fermentation, could be regarded as 'sweet' beer or a cool beverage).Only 8% of female heads reported that they occasionally threw away smallportions of sadza and relish. Usually cooked food was put into a bowl orlarge plate from which members of the household shared the meal. Thosewho threw away left-over portions did so for health reasons, arguing thatthese portions were not good for recycling. Only one respondent gave left-over food to co-residential persons.My respondents generally kept no domestic pets. Only one widowedhome-owner kept a dog on her premises, for security reasons. Mostpremises were not fenced, so combined with the women's mobile lifestyle,the absence of domestic animals was quite understandable. This explainedwhy most female heads were not aware of the 'dog tie order' by theMasvingo Town Council during the first two weeks of July 1995, when anystray dog was shot on sight.The absence of pets could also be related to sentiments mooted bysome respondents during conversations. They felt that they could notafford to keep pets as this would negatively affect their meagre householdbudgets. They viewed pets as additional consumers. Keeping a pet certainlywould have raised domestic expenditure and the organisation of mealsand leftovers would have been different if the respondents had had pets.158 SOURCING AND SELLING FOOD IN MASVINGOThe organisation of meals was also interesting in terms of who wasresponsible for the cooking and distribution of prepared meal portions.Other than the occasional invitations of a friend, neighbour or relative,there were virtually no shared meals. Meals were usually taken withhousehold members only, in those cases where households had morethan one person.SELLING FOODCouncil-designated vending areasAccording to the Masvingo Municipality, all food vendors had to registerand pay a monthly licence fee. Every food vendor was supposed to operatefrom a Council-designated site. At the beginning of 1995, Council was stilltrying to control the number of vendors operating in the city, but thispolicy seemed to have been abandoned by the end of that year. By thenCouncil was only interested in collecting registration and monthly licencefees, not in attempting to control the number of food vendors. However,the requirement that vendors operate from a designated site remained inforce. At the designated vending sites, food vendors were only allowed tosell perishable food items, mostly fresh produce, a practice defied bymany women vendors. Non-perishables such as mealie-meal and sweetswere considered by council to be illegal tradeables, despite the vendors'payment of the monthly vending fees.Sixteen of the 19 food vendors in the study were licensed. Twelveoperated from Masvingo town centre, four from Mucheke markets andthree from Rujeko markets. At the beginning of 1995 licence fees were $21per month, but were increased to $27 from August 1995. The municipalityenforced payments by registered vendors by disconnecting water supplies,which sometimes happened long after the lodger had moved to someother premises. The practice of water disconnection was still in force atthe end of 1995.A minority of vendors, especially those operating in the townships,operated from the licensed Council roofed markets. Each vendor wasallocated a permanent operating area in return for the monthly vendingfee. There was no such formal shelter in the town centre where themajority of food vendors in my sample operated. Town-licensed foodvendors were supposed to operate from open-air markets, such as theone opposite Aroma bakery. At the open air markets the women wereexposed to the scorching sun during the dry season and the winds andrains during the wet season, yet they were required to pay the sameamount of money in licence fees as those who operated from formalCouncil shelters.In November 1995, shortly before completing my fieldwork, the areafacing Aroma Bakery, the space previously allocated to the women in myV. N. MUZVIDZIWA 159study, was designated a site for all vending operations in town. Othervending operations relocated to this site included those selling non-fooditems such as watches, toys, doilies, art and craft products. The immediateresult for the food vendors was increased market opportunities. Salesshot up drastically for the four women who operated permanently at thesite, and another four women started spending more time at the site. Itappeared that this site had become more lucrative for food sellers due tothe increased volume of traffic in a more diversified market place.ILLEGAL VENDING AREASThree illegal food vending areas were pavements, bus terminuses andverandahs. Although Masvingo Town Council saw food vendors assedentary, in actual practice most food vendors were semi-mobile andsome were fully mobile without any base. In town as well as the townships,it was common to see food vendors walking on foot and chatting to would-be customers trying to persuade them to buy their wares. This practicewas common the week before teachers' pay day when sales were downdue to low demand because people had run out of money,Although the vendors thought of particular spots as 'their' operationalareas, from time to time they moved and set themselves up at other spotsalong pavements they considered strategic to their business operations,mostly close to places that attracted large numbers of people. The foodvendors tended to sit in groups based on friendship networks. Pavementfood vending was mostly characterised by order rather than chaos, eventhough the municipal regulations tended to criminalise vending operations.During the course of fieldwork, for over seven months I observed atthe Mucheke long-distance bus terminus, illegal food vending operationsinvolving largely women and children. The majority of the kids were ofschool-going age. The market for terminus food vendors was the travellingpublic. Food vending at the terminus was characterised by a very highdegree of uncertainty, with the police harassing and threatening to arrestthe vendors. The vendors were always ready to take to their heels.The long-distance terminus, unlike vending sites in town, attractedmany young boys, youngsters aged between 13 and 25 who were alwayshovering and milling around particularly at peak times. A group of youngpickpockets, numbering about a dozen, sometimes acted as loaders orjaggers (touts) pretending to help travellers. A group of six beggars,operating in pairs, frequented the terminus. This is the environment underwhich some of my respondents operated, yet according to my observationsand discussions with them, none of the women food vendors lost anywares to the terminus 'boys'. Women food vendors had only one fear atthe terminus, the fear of harassment, arrest and confiscation of wares bythe police, not by the young crooks known as tsotsis.160 SOURCING AND SELLING FOOD IN MASVINGOVerandah trading in Masvingo existed on a very low level, unlikeObbo's (1982, 137) description of Kampala where for every three housesthere was a woman selling something, no matter how small, from herdoorstep. In Masvingo, due to the rigorous enforcement of vendingregulations and the desire on the part of the Municipality to maximisefinancial gains from licensing, there were very few food vendors operatingfrom their doorsteps. These few who practised it risked being fined by thepolice.Food vending constraintsWomen vendors saw the ever-present threat of police arrest andharassment as the single biggest obstacle to their businesses. Most of mysample vendors, except those who operated from Rujeko, had experiencedofficial harassment in the 12 months preceding the study. Women vendorslived in constant fear of official harassment. Even though most were legaloperators who paid a monthly fee to Council, they were usually arrestedfor failing to observe the rule that they operate from designated points.Harassment came from both the Zimbabwe Republic Police (ZRP) and themunicipal police.Some vendors in my sample had been fined three times in less thanthree weeks during April 1995. Illegal vending carried a fine of $30 eachtime, and goods were confiscated to act as a further deterrent. One of thevendors in my sample subsequently moved out of the premises where shehad been living, and I never saw her again. Other vendors said she had leftfor the communal areas, because the fines had destroyed her vendingoperations. The female vendors in the study wanted harassment andarrests to stop, and the municipal bye-laws and regulations to be relaxedso as to allow mobile vending. The women were not against licence fees,but they wanted a reduction in fees commensurate with the level ofservices Council offered. For instance, the food vendors in town operatedin the open, and depended on trees for shade. During storms and strongwinds they took cover under nearby shop verandahs, yet they wererequired to pay the same amount in fees as those who were provided withmarket stalls. Council justified their payment of these fees by arguing thatthe town women used its water and toilet facilities in the course of theiroperations.During 1995, a number of police raids were targeted at women foodvendors. Officials wanted to rid the city of illegal vending, and in theprocess publicised vendors as people who posed a health risk to cityresidents. Yet the sites where vendors operated were kept clean by thevendors. Food vendors in Masvingo town centre made it a point at the endof each day to remove any rubbish in the vicinity of their operational spotincluding that from shoppers in town, who had nothing to do with theV. N. MUZVIDZIWA 161vendors. In many ways the vendors contributed to the general cleanlinessof the town, and contributed an unrecognised subsidy to the generalmaintenance costs of the municipality.One food vendor thought that police actions and harassment wenttoo far during the April 1995 raids. She noted that policing licenses wasunderstandable, but not when it got to the point of threatening the women'svery source of living. When police chase women vendors and treat themas worse than thieves then there is something wrong. Women are simplytrying to support their families. Police must be more sensitive to people'sneeds. Police spend too much energy chasing poor women instead ofthieves and other undesirables. She was even more suspicious of whathappened to confiscated goods.I have suspicions that the lorry loads of tomatoes, vegetables, bananasand other food items they confiscate are either eaten or given to friendsor maybe resold through some market women. It pains us a lot to thinkabout these possibilities.The state's response to Masvingo female vendors reflected a narrowly-conceived, male-dominated official policy of 'development' thatmarginalised the poverty alleviation strategies of informal sector operatorssuch as the women food vendors. The officials' concern was with orderlydevelopment and a clean city, since Masvingo is a leading tourist town byvirtue of its proximity to the Great Zimbabwe Monument. In October andearly November 1995, all the officials who addressed meetings with vendorsorganised to relocate them were men, representing different organs of thestate and Municipality, such as the Zimbabwe Republic Police (ZRP) andMasvingo Municipal Health and Housing and Community Servicesdepartments. The Vendors' Association representatives, ironically, werealso all men. The power conflicts that women encountered in the course ofvending reflected the orchestrated exercise of male power in society,using idioms of male-female behaviour patterns perceived as appropriateby men. The masculine state tried to control and to undermine any gainswomen vendors had made in achieving an independent existence outsidemale control. Women vendors were not simply faced with a hostile marketenvironment, but had to find ways and means of dealing with hostile statepolicies and functionaries as well as local-level restrictive regulations intheir quest to survive in the city.Thus the public meetings became a mere re-enactment of genderroles, demonstrating how women must first negotiate a space for acceptablepublic behaviour. Women vendors behaved in a manner that did notthreaten male control. They ululated and nodded their heads in agreementwith official suggestions; but they then continued, contrary to male, officialexpectations, to sell on the pavements. The women had to subvert someof the regulations without being seen to be doing so. The vendors adopted162 SOURCING AND SELLING FOOD IN MASVINGOsvejkist strategies in their dealings with officials, giving the impressionthat they were going along with male officials' suggestions when they werenot. It was the women's svejkist stance that enabled many to continue tohang on to town life. Throughout 1995, despite the arrays of municipalregulations that restricted food vendors, my respondents operated as ifthe rules were not there.FOOD VENDORSAll 19 full-time food vendors in my sample operated as sole traders. Theywere responsible for decisions that affected their vending enterprises.The study's food vendors fell into the 'hanging-on' and 'coping' categories:only one food vendor was climbing out of poverty. In their day-to-dayoperations, six food vendors got assistance from their own children orrelations or from hired labour. Three food vendors employed salesassistants. For instance, respondent V employed two assistants to whomshe paid $60 each per month. She provided them with food rations andshared her premises with them until she entered a mapoto union.The women had varied food vending experiences, ranging from lessthan 12 months to 23 years (Table 8), but 50% of food vendors hadoperated their food vending businesses for less than three years. ESAPresulted in the sudden upsurge in food vending operations. With thederegulation of economic activities and relaxation of previous restrictionson food vending, more women went selling food in the streets to feed theirhouseholds.Table 8RESPONDENTS' FOOD VENDING EXPERIENCESFood Vending Experience Number of Food VendorsLees than a year 31 to 3 years 64 to 6 years 27 to 10 years 4More than 10 years 4Total Respondents . 19The main reason identified by women in the sample for selling foodwas the desire to earn an income. Given the women's level of educationand lack of employment opportunities, vending offered a chance to earn aliving without a large capital outlay. Eighty-four per cent of food vendorsV. N. MUZVIDZIWA 163had a primary level education or less. For those who started vending whenthey were still married, it allowed them to operate from their homes in amanner acceptable to their husbands.Food vending required physical mobility to source and sell the wares,whether operating in the city or in the residential areas. Some of the foodvendors went 300 km out of Masvingo to buy foodstuffs. Those whooperated from Council-designated selling points moved in search of waresand markets. A common joke among the female vendors was Ukanyarahaugwazi (If you are shy you will lose out; If you get tired you will notachieve your objective). For instance, the pavement food vendors inMasvingo left the bulk of their commodities at the legal vending site inorder to minimise losses due to confiscation of wares during the frequentpolice raids. Hence the women spent considerable time moving back andforth between the Council's legally-designated vending site and their illegalpavement sites.Successful food vending meant the adoption of multiple vendingstrategies in order to maximise returns. Food vendors operated fromdesignated sites (eg. Council markets), but also from pavements, streets,bus terminuses and private verandahs. Generally food vendors sold onlyfor cash, but in some cases a known customer could buy on credit. Whilereturns on food vending were quite low, cash sales ensured a daily flow ofincome.The food vendors' initial 'capital' came from three sources: savings,loans and gifts (Table 9).One needed only a small amount of money to start operating as a foodvendor: in 1995 less than $50.FOOD VENDORS'Own SavingsLoansMonetary giftTotalTableSOURCE9OFINITIAL CAPITALTotal99119%47476100The food vendors strongly desired to retain control of their ownbusinesses by avoiding credit. They felt that it was easy to lose one'sbusiness to money lenders through indebtedness, and had numerousexamples of people who had lost their businesses due to their inability torepay a loan. Only two in the study had at some time in the past obtained164 SOURCING AND SELLING FOOD IN MASVINGOcredit from sources other than personal ones. To them, taking a loan waslike tying a rope around one's neck. Given the unpredictable businessenvironment under ESAP, the vendors' responses were quite rational.They preferred to turn to marounds and friends for assistance rather thanfinancial institutions, perhaps because only seven of the 19 had a bankaccount.Food vendors were an extremely busy group, 12 operated six days aweek from Monday to Saturday. Sundays were spent doing householdchores (laundry, tidying up). Some attended church, others visited friendsor kin, yet others used Sundays to replenish supplies. The women hardlyknew any leisure. All, without exception, worked for more than 13 hours aday. Vendors started their working day around 5 am, buying at the producemarket in the early hours so that by the time customers woke up theycould serve them. Competition was for the best produce in the market andthose served first tended to get the best fresh produce. This was useful inattracting customers. Food vendors went to bed late, around 10 pm, inmany instances because they also did part-time knitting and crocheting ona commissioned basis for cross-border traders.Despite the obstacles faced by female food vendors, 14 were certainthat their food vending operations would sustain their households for aconsiderable time to come, only three saw a bleak future, and two werenon-committal in their attitude towards the future. Against the view thatthings would improve in the near future and guarantee the viability oftheir businesses, three pointed out that vending was not lucrative andthat success was too dependent on the ability of the operator to conductbusiness. If one fell ill, that could spell the end of one's business. Mostvendors had no intention to expand their businesses. They just wanted tomaintain current operational levels that they could manage effectively.Ten food vendors had made savings, whilst the rest lived on a hand-to-mouth basis.FOOD VENDING AS A SOURCE OF INCOMEOf the 19 food vendors, nine had no other source of income, 10 did part-time knitting as well as crocheting for cross-border traders. The monthlyreturns for this put-out work were very small, generally less than $100.However, this additional source of income was appreciated. Womenvendors knitted at slack times during their vending operations, and onewoman, in addition to food vending, also sold second-hand clothes sourcedfrom Beitbridge. Food vending was particularly slack in the early morninghours, from 6 am to 10 am.It is also worth noting that 18% of my Masvingo sample derived asecondary source of income from part-time food vending activities, realisingV. N. MUZVIDZIWA 165a monthly income of less than $100. At the time of research, food vendingwas for all practical purposes a survival strategy and nothing more.Income from food vending produce: Fresh fruits and vegetablesFood vendors generally specialised in tomatoes and fruits. In many waystomatoes were treated as more than just a vegetable. In the study only twofood vendors, a divorcee and a widow, specialised in the sale of grain andsweets respectively. In Masvingo, women food vendors were referred toas vakadzi vemadomasi (tomato women), because virtually all soldtomatoes. There are many reasons why tomatoes are of strategicimportance in the fresh produce market. Tomatoes functioned as relish oflast resort and were used with nearly every other type of relish. Thisresulted in some kind of inelasticity in the demand for tomatoes. In hotclimates tomatoes tend to last longer than some green-leaf vegetables,although they are very susceptible to fungal infection. Despite the factthat tomatoes were in high demand, correspondingly high prices were notobserved.The food vendors bought their tomatoes (and other vegetables) froma variety of sources both within and outside Masvingo, from farms andirrigation schemes as far away as Esigodini (nearly 250 kms west ofMasvingo) and Birchenough Bridge (160 kms east of Masvingo), and fromthe small peri-urban plots around Masvingo. Supplies bought from outsidethe city cost less than half the Masvingo price. For instance, in February1995 a carton of tomatoes cost $15 from farms in Esigodini compared to$35 in the wholesale market in Masvingo. Usually the women purchased inbulk at source, a minimum of ten cartons or crates when they orderedoutside Masvingo town. Some women travelled for orders on a weeklybasis, while others made monthly trips to centres of supply. A commonpractice amongst the female vendors who depended on external markets,was to minimise transport costs by teaming up into groups of three to fivewomen, each of whom went out for orders in rotation. Each time a womanbrought in supplies, she sub-divided the lot with two to four other women.Of the 19 food vendors, five operated individually including the one womanclimbing out of poverty. The other 14 operated in three groups of three,with one group of five members.In all cases the women in such groups contributed towards thepurchase of the fresh produce and sometimes but not always assistedwith transport costs for the purchases. Orders were brought back bybuses. The standard freight cost per basket was $5 in 1995 irrespective ofdistance. In a single trip the women could carry up to ten baskets. Thedegree to which the women used networks based on friendship for businesspurposes was quite impressive and showed a high degree of organisation.For example, members of the group of five noted above took turns going166 SOURCING AND SELLING FOOD IN MASVINGOout to source orders. They contributed towards the cost of transport andorders. Thrice a week three members of this group went out sourcingorders.As noted earlier, very few women food vendors had surplus vegetablesfor sale from their gardens. Especially during the rains, some procuredtheir tomatoes and vegetables, as well as fruit supplies, directly from ruraltraders. In some cases, village supplies were coordinated through kinnetworks. Women vendors also purchased fresh produce from urbanmarkets, especially wholesalers, supermarkets and fresh produce shopsin Mucheke and in the town centre from places like Tanaiwa and Vitagreen.Nearly half of the women replenished supplies from delivery lorries thatbrought fresh produce from production sources. The lorry deliveriesspecialised in tomatoes, vegetables (especially cabbages) and fruits. Lorrydeliveries were from as far away as Mutare, some 300 kms north-east ofMasvingo. The lorry park was in Mucheke, close to the long-distance busterminus.Income from non-perishablesThe percentage of food vendors who traded in relatively non-perishablefood items varied greatly throughout the year. During the harvest season(around February to May), up to half of the women sold some non-perishable foods, including dried maize, groundnuts, roundnuts, cow peas,mufushwa (dried okra) and sweet potatoes. Most of these foods were soldthroughout the year, by about a quarter of the women vendors. The mainsource of supply of the non-perishables were village producers, farms andirrigation schemes.In March 1995, one woman started selling mealie-meal and peanutbutter, having borrowed her start-up capital of $100 from a maternal aunt.She operated from an open space near Mucheke long-distance bus terminusas well as from her residence, although as a registered vendor she wassupposed to operate from a Council-designated selling point. Where shesold was considered illegal, even though she paid her monthly vendinglicence. She also operated seven days a week. Despite increases (from $18per bucket in February 1995 to $35 in October), the demand for mealie-meal proved inelastic. Until the end of October 1995 she sourced hersupplies from rural traders who came to town to sell their wares. She thenbecame dependent on the Grain Marketing Board for her supplies, andstopped selling peanut butter, as she was no longer getting regular suppliesof peanut butter from rural traders. She compensated for this by increasingher hammer-milled maize-meal sales.She packaged the mealie-meal into one-kilogramme bags, convenientfor her customers' needs. Commercially-packaged mealie-meal came inminimum units of five kilogrammes. Her hammer-milled meal was sold atV. N. MUZVIDZIWA 167about half the price of the industrially-milled product. A study conductedby the Ministry of Public Service, Labour and Social Welfare in collaborationwith the United Nations Children's Emergency Fund (UNICEF), reportedthat 5 kg of roller meal was selling at $10 whilst 5 kg of hammer-milledmeal was going at $5.50. My respondent was selling 1 kg at $1.25 (this hadrisen to $1.75 by October) and yet demand outstripped her supply. Shemet the needs of small, one-or-two person households without the moneyto buy in bulk, who purchased just enough to make a couple of meals.Income from other produce: PerishablesFood vending in perishables other than fruits and vegetables was verylimited among my Masvingo sample. One food vendor had been sellingcooked food to construction workers in town since 1984. She receivedsome assistance in the preparation of the food from her brother's daughterwho stayed with her, but did the selling herself. She had standing ordersfrom the construction workers.Other perishables sold by a few vendors on an ad hoc basis includedfreezits, eggs, milk and chickens. During the rainy season (mostly duringthe period January to March 1995), a number of women sold mushrooms,which had been brought in by rural traders. Mushrooms are highlyperishable, but due to scarcity they sold like hot cakes as a seasonalproduct.FORMAL AND INFORMAL MARKETSThe food vendors were linked to commercial as well as small-scaleproducers. They knew where and when certain goods were available,sourcing supplies directly from producers. Informal trade dominated foodvending in Masvingo. Presumably through such informal marketingarrangements, both producers and vendors benefited as they weresometimes able to circumvent state sales tax, thereby reducing the costsof fresh produce to consumers, and the cost of urban subsistence. Manyhouseholds would not have been able to subsist at the level at which theydid, had it not been due to the hard work of these women. In many waysthe food items sold by food vendors subsidised the cost of subsistence inthe town. Fresh produce would have been much more expensive if thewomen had not sourced directly from producers.Food vending was a very demanding job. It was through the efforts ofthese and other female food vendors that fresh produce supplies werekept flowing into cities like Masvingo. Not only were the urban residentssupplied with fresh produce, but some rural residents were also suppliedby food vendors in my sample. Through observations carried out during1995, as well as discussions with the women vendors, it became clear that168 SOURCING AND SELLING FOOD IN MASVINGOtheir market extended well beyond the confines of Masvingo town. Of my19 vendors, only three catered exclusively for the urban market, while therest served both urban and rural clients. At month-ends, the food vendorsstocked up on fresh produce in order to take advantage of rural publicservants, who came to Masvingo for banking and shopping. These ruralformal-sector employees of government and private companies purchasedtomatoes and other fresh produce items in town.Vending as a market activity required a lot of resourcefulness andenergy on the part of the female heads of household as vendors. Officialconstraints imposed on the women posed a greater risk to the success oftheir operations than any other factor. Although vending was not thateasy for many, it was all that was there to depend on for a living. InMasvingo vendors were an integral part of the city; they interacted withformal sector operators in many and diverse ways. They were a vital cogin the fresh food distributive network. There was evidence that, due to thepresence of food vendors, formal businesses close to vending sites attractedcustomers who could combine a visit to the shop with buying food.Masvingo food vendors complemented the existing array of goods suppliedby retailers and in addition created a shopping environment that attractedshoppers to the street. Vendors transformed pathways into shoppingmalls, resulting in increased business to both established businesses andvendors. Food vendors operated in both the formal and informal markets.VENDING AND THE POVERTY TRAPFood vending returns did not enable many to invest and climb out ofpoverty. The role of food vending dates back to the colonial period whenit was introduced in urban centres as a welfare cushion mainly for thosehouseholds with no male bread winner. Food vending was never expectedto lift the women out of poverty, but rather to prevent them from goinginto prostitution.Food vending was a hanging-on strategy. For food vendors, illness orpolice harassment and confiscation of goods could result in downwardmobility into the 'burnt-out' category and moving back to a rural village.However, the question to ask is: did the one case who used food vendingto climb out of poverty, indicate conditions under which food vendingcould be turned into an investment strategy?The successful vendor respondent V operated from two sites inMasvingo town, one an illegal pavement site just outside Balmainsupermarket and the other at the legally-designated site opposite Aromabakery. She left the bulk of her day's orders at the legally designated siteand carried only a small quantity, so that if confiscated by the police theloss would be small. In addition she had at any one time an assistant whooperated as a mobile vendor. Throughout fieldwork she had an assistantV. N. MUZVIDZIWA 169who moved around town with an even smaller amount of wares. Althoughshe was a legally registered vendor, she was not immune to arrest sinceanybody who operated outside the designated zone was deemed an illegaloperator.V ordered bananas twice a week from a Mutare delivery truck. InOctober 1995, she noted that the profit margin for a carton of bananas hadgone down from at least $30 to $15. This was due to two factors: thenumber of bananas in a carton had been reduced from 100 to 85 or so, andby August 1995 the price of the carton had gone up to $45 from $30 at thebeginning of 1995. The purchase price for her customers of a singlebanana had gone up from $0.60 in early 1995 to $0.70 in October 1995,giving her a gross margin of roughly 30%. V also sold tomatoes andvegetables, and most perishables and non-perishables noted above. Shesourced scarce goods such as mushrooms when they were selling at peakprices. She had operated in a three-person food sourcing group, but at thetime of fieldwork was doing her ordering all by herself from places as faras Nyanyadzi, Esigodini and Chivi. She knew most of her customers byname and related to them in the idiom of kin relations. She was highlymobile, since her eight-year-old twin daughters were staying in the villagewith their maternal grandmother. She had a split household and onlythree dependents. This climbing-out-of-poverty vendor had registered onthe Council's housing waiting list and was quite optimistic that soonerrather than later she would get a stand on which to build a house. It wasnot possible in a study like this to be definitive as to why she was successful.It appeared that a high degree of mobility and shrewdness were neededfor success. She was also one of the few who concentrated all her energieson vending. The majority of food vendors remained in the 'hanging-on'group, engaged in multiple income-earning activities, and were double-rooted (84% of all food vendors were double-rooted).Vending was generally not a strategy to climb out of poverty, but a'hanging-on', and to a lesser extent a coping strategy that enabled manyurban female heads of households to stay in the city for as long aspossible. Through vending operations my respondents noted that theywere able to get their children food, clothes, at times shelter, school feesand hospital fees. Some food vendors were gaining a new confidence thatenabled them to defy Council regulations on the use of vending space.Most food vendors noted that they had gained autonomy and independencethrough their incomes, which might perhaps translate itself into collectiveaction in the future.VENDING PROSPECTSDuring the year I observed an increase in the number of food vendorsoperating in town and in the townships. Yet, as noted above, 74% of food170 SOURCING AND SELLING FOOD IN MASVINGOvendors in my sample were in the 'hanging-on' category. For all but oneexceptional woman, food vending was nothing more than a 'hanging on' or'coping' strategy. Food vendors had very limited financial means. Lack ofcapital and unwillingness to raise capital through formal credit meantmany food vendors remained at subsistence-level operations.The food vendors' attitudes reflected a pervasive ideological mind-setagainst credit which was widespread in Zimbabwe (except for those whosee credit as free money, mostly the ruling elite and the favoured few).Linked to capital were the problems related to difficulties in sourcingorders. The few successful women vendors were those who could travellong distances in search of fresh produce, and employ others to sell whatthey sourced.Lack of capital, increased competition, difficulties in sourcing orders,as well as official harassment, all worked to limit the profitability of foodvending operations. Whilst 26% of food vendors in the study showed thatfood vending could be more than a 'hanging on' strategy, and one hadeven climbed out of poverty, the generally low returns from vending madeit difficult if not impossible for most food vendors to move out of thepoverty trap. Policies of 'encouragement' (byway of increasing the numberof licenses) had resulted in overtrading and a consequent decrease inprofitability, which appeared to threaten further the viability of foodvending operations. With more people entering vending, it was becomingdifficult to secure orders cheaply from town suppliers. Sourcing goodsdirectly from producers required transportation money which the majorityof the vendors did not have.Three other constraints related to labour, storage and transport. Thefood vendors operated low-return solo businesses. Given the level ofreturns, most food vendors could not hire extra labour, yet the mostsuccessful food vendor did.Regarding storage, all food vendors kept their wares at their places ofresidence, generally under a bed or tucked in a corner. Because the foodvendors used the rooms where they lived as storage places, it meant thatthey were limited in the amount of wares they could store at any one time.However, some food vendors operating in town had befriended peoplewho rented accommodation in town and left their unsold wares with their'town' friends for safe keeping overnight.In order to minimise transport costs, food vendors carried their goodson their heads most of the time. They simply walked whenever andwherever they went within Masvingo town. Paid transport was reservedfor journeys outside Masvingo.CONCLUSIONThis article has shown how Masvingo respondents employed differentstrategies to ensure that food was available for their households, throughV. N. MUZVIDZIWA 171shrewd shopping patterns as well as the strategic use of the householdgarden. The organisation of meals and household budgeting demonstratedthe ability of the women to manage with very little. Urban food vendorswere strategically positioned to benefit from both the urban and ruralmarkets. Vending was necessarily a 'political' activity, since food vendorshad to claim public space for their vending operations in the face ofofficial harassment and opposition. Policies related to illegal and designatedvending areas limited the development of food vending in Masvingo.Vending problems were largely structural, including the free marketcompetition that posed practical dilemmas to the food vendors concerningfood vending viability and profitability. Increased competition wasproducing an over-traded food market, and this had resulted in depressedreturns for food vendors, judging by their comments.BibliographyBailey, F. G. (1993) The Kingdom of Individuals: An Essay on Self-Respectand Social Obligation (Ithaca, Cornell University Press).Central Statistics Office (1993) Census 1992: Provincial Profile, Masvingo(Harare, CSO).Hansen, R. T. (1980) 'The urban informal sector as a development issue:Poor women and work in Lusaka, Zambia', Urban Anthropology, IX, (ii),199-225.Kemper, R.V. (1981) 'Obstacles and opportunities: Household economicsof Tzintzuntean migrants in Mexico City', Urban Anthropology, X, (iii),211-229.Logan, K. (1981) 'Getting by with less: Economic strategies of lower incomehouseholds in Guadalajara', Urban Anthropology, X, (iii), 231-246.Matarise, T. (17 Feb., 1995) 'More poorer folk turn to mugayiwa', TheHerald, C12.Muzvidziwa, V. N. (1997) 'Rural-urban linkages: Masvingo's double-rootedfemale heads of households', Zambezia, XXIV, (ii).Obbo, C. (1982) African Women (London, ZED Press).