maasai socioeconomic conditions: a cross-border comparison

P1: GVM/GDX

Human Ecology [huec] pp417-huec-368385 February 28, 2002 9:38 Style file version Nov. 19th, 1999

Human Ecology, Vol. 30, No. 1, March 2002 ( C© 2002)

Maasai Socioeconomic Conditions:A Cross-Border Comparison

Ernestina Coast1

A single round household survey was conducted between October 1997 andMay 1998. Information was collected on 1545 Maasai households in Kenyaand Tanzania using a standardized questionnaire. These data represent the firstlarge-scale, cross-border comparable survey of socioeconomic indicators forthe Maasai. The structure of the study is similar to that of a natural experiment:one ethnic group living in two very different nation-states. The survey resultsdescribe a current “snapshot” of Maasai livelihoods. Living arrangements aredescribed at both the household (olmarei) and multihousehold (enkang) level,together with changes in housing type. The roles of cultivation, transhumantmigration, and employment are described, including a consideration of theprevalence of murranism (warriorhood). The implications of rising levels ofsedentarization and cash crop production for Maasai economic diversificationare explored. Participation by Maasai in the tourism industry is extremely lowdespite the proximity of Maasai to major international tourist destinations inEast Africa. Levels of formal education show a marked sex bias in completedprimary school education in both countries, with far fewer women than menhaving attended school.

KEY WORDS: Maasai; Ngorongoro Conservation Area; Kenya; agropastoralism.

INTRODUCTION

The Maasai of Kenya and Tanzania are one of the best-known pastoral-ist populations in the world; indeed Spear suggests “Everyone ‘knows’ theMaasai” (1993, p. 1). The economic and social conditions of the Maasai havechanged throughout their history in response to a myriad of factors operating

1Department of Social Policy, London School of Economics, Houghton Street, London WC2A2AE, United Kingdom.

79

0300-7839/02/0300-0079/0 C© 2002 Plenum Publishing Corporation

P1: GVM/GDX


80 Coast

over a variety of spatial and temporal scales. In recent decades the influenceof nation-states, monetization of the traditional economy, formal education,land tenure changes, and demographic factors have all played a part in shap-ing the current socioeconomic situation of Maasai in Kenya and Tanzania.

As neighboring nation-states, Kenya and Tanzania lend themselvesreadily to studies that “compare and contrast.” The similarities are striking:a common climate; similar natural resource bases; rapidly growing popula-tions of small subsistence farmers; common cultural groupings; and historicalBritish colonial rule. The differences are equally remarkable. Postindepen-dence Kenya has (until very recently) pursued policies that emphasize eco-nomic growth over equity, private sector development, receptivity to foreignprivate investment, preservation and extension of colonial institutions, andthe maintenance of close ties with Britain. Tanzania, on the other hand, hasbeen typified by socialism, self-reliance, a more equitable distribution of thecountry’s wealth, state intervention in and ownership of the economy, a re-duction in reliance on agricultural exports, and forced villagization (ujamaa)of the rural population (from Barkan, 1994).

The national level divergence in policies between socialism in Tanzaniaand capitalism in Kenya are highlighted within Maasailand. In Kenya, theimpact of the Group Ranch programme has had profound implications forMaasai access to land. The Group Ranch concept was rooted in postindepen-dence Kenya and is based on the premise that if livestock production couldbe raised, bringing it into the commercial sector, then any wealth generatedcould be ploughed back into development of semiarid areas (Evangelou,1985). Changing land tenure toward individual or group ownership was seenas key in the commercialization of pastoral production systems. Initially, thepolicy involved the registration of small (10 km2) areas of land to individualMaasai. However, as Homewood points out “these were neither ecologicallynor economically viable for livestock production” (1995, p. 338). The naturalsuccessor to the original plan was to parcel up land into larger blocks, to beheld by cooperatives of herd owners. Recently, however, the Group Ranchsystem has been subject to subdivision of land in a move toward individualland ownership in line with Kenyan government policies (Grandin, 1988;Homewood, 1995; Rutten, 1992).

The development of ranching associations in Tanzania took place at thesame time as the group ranches in Kenya. However, that is where the similar-ity of these policies ends. Each ranch had a coverage of thousands of squarekilometres of previously communal land and the programme was carried outwith large overseas financial and technical assistance under the auspices ofthe Tanzanian government. In spite of laudable rhetoric to integrate Maasaipractices and objectives, the project involved “little or no consultation withMaasai on geographical boundaries or specific aspects of development”

P1: GVM/GDX


Maasai Socioeconomic Conditions 81

(Homewood and Rodgers, 1991). The result was that only one ranching as-sociation achieved the aim of ensuring title to the land. Whereas the groupranches in Kenya have had a lasting impact on the tenurial composition ofKenyan Maasailand, in Tanzania the net result was viewed by the Maasai as“just one more case in a long history of broken promises” (ibid).

The Tanzanian data for this study were collected in the NgorongoroConservation Area (NCA). NCA occupies a unique position in Tanzania;indeed “other laws governing the lives of Tanzanian people are subsidiaryto those of the NCA” (Homewood and Rodgers, 1991). The history and ob-jectives of the NCA have been well documented elsewhere (Arhem, 1985;Thompson, 1997). NCA represents a unique experiment in joint land use,between conservation, tourism, and agropastoralism. The balance betweenhuman occupation and wildlife concerns are continually being reassessed,the result being increasing levels of tension and potential conflict betweenwildlife managers and resident Maasai. Currently of concern, for example,is the issue of whether or not cultivation should be permitted within theboundaries of the NCA. An uneasy truce exists at the moment, with small-scale subsistence agriculture allowed, provided that ploughs are not used.Issues in NCA extend beyond its boundaries, with interested parties includ-ing the Ngorongoro Conservation Area Authority (NCAA), local NGOs,national government, international agencies, and tourism. These thumbnailsketches demonstrate the heterogeneous context of Maasai, and the widersocioeconomic setting within which most Maasai live.

The design of this study can be thought of as a natural experiment. Thatis, in order to conduct a representative survey of the Maasai, it was neces-sary to collect data in both Kenya and Tanzania. A situation therefore arisesof one major controlling factor (ethnicity) whilst varying other factors, mostnotably nation-state. The use of the term “experiment” is perhaps somethingof a misnomer in the social sciences: it implies total control of dependent andindependent variables. In a social science setting, however, there are manyconfounding factors and perfect “control” is absent. This caveat notwith-standing, given that the opportunity arose for an approximation to a naturalexperiment design, it is still a useful—if inexact—framework for analysis. It isan approach worth pursuing because, to echo the words of Hill and Randall,“Natural experiments in the human sphere are not all that common and thechance to look at the behaviour of the same ethnic group living in differentcircumstances . . . is too valuable to waste” (1985, p. 39).

This study describes the results of a large-scale (n = 1545) householdsurvey of Maasai2 in Kenya and Tanzania. The data were collected using a

2Here, Maasai refers to the broadest conceptualization of “Maasai” as an ethnic group,using Hutchinson and Smith’s (1996) definition of an ethnic group. The debate surrounding

P1: GVM/GDX


82 Coast

standardized questionnaire adapted to a Maasai context and represent thefirst cross-border directly comparable set of data on current Maasai socio-economic conditions. The strength of the survey data presented here is theircomparability and size. The major limitation is the shallowness of informa-tion that has been gathered using a survey. The purpose of this descrip-tion is to complement more detailed, focused studies of Maasai livelihoods(e.g., Bekure et al., 1991; Galaty, 1981; Helmut, 1995; Ibrahim and Ibrahim,1995), not to replace them. The data must be contextualized within the widerproject3 within which they were collected. The project was multidisciplinary,concerned with the long-term outcomes of different land-use policies on en-vironment, wildlife, and socioeconomic indicators in the Serengeti–Maraecosystem in Kenya and Tanzania.

As highlighted by Spear “‘Maasai’ and ‘pastoralism’ have become soclosely linked in the historical and ethnographic literature . . . that Maasai arecommonly viewed as prototypical pastoralists” (1993, p. 2). To read literaturerelating to the Maasai, one would be left with an unclear idea of how theycurrently derive their living. Maasai are variously referred to as pastoralists(Grandin, 1991; Hedlund, 1980; Holland, 1987; Jacobs, 1979; Sindiga, 1992;Talle, 1994), “pastoralists (with) some degree of mobility” (Galaty, 1992), a“specialised pastoral community” (Bonte and Galaty, 1992), “nomadic pas-toralists” (Kipuri, 1998), as including “sedentary agriculturists . . . [and] . . .nomadic pastoralists” (Asiema and Situma, 1994). Diversification within tra-ditional Maasai economies, particularly in Kenya, has been subject to det-ailed studies (e.g., Rutten, 1992). The aim of this study is therefore to broadenour understanding of contemporary aspects of Maasai livelihoods and life-style in both Kenya and Tanzania using a comparative analysis of survey data.

The survey results provide a current “snapshot” of Maasai livelihoods.Data were collected on a wide range of socioeconomic variables, chosenspecifically to capture both the changing and enduring aspects of Maasailivelihoods and human capital.4 These variables may be divided into four

“Maasainess” (Spear and Waller, 1993) is noted, and implicit throughout the study. Sectionaldifferentiation was not incorporated into the study as members of different traditional unitslive mixed together in the geographical areas of the study, e.g., Salei and Ngorongoro Maasaiin NCA.

3“Impacts of land-use policy on environment, wildlife, demography, and socioeconomic indica-tors in east African savannas: the Serengeti Ecological Unit,” funded by the European UnionGrant No. IC18-CT96-0070.

4It is acknowledged that livestock ownership is the measure of wealth most commonly usedby Maasai themselves. However, data on livestock ownership were not collected. Reliablelivestock data are extremely difficult to collect in a large-scale survey, and it became apparentduring the pilot study that the collection of household livestock ownership would jeopardizethe entire household survey. Intrinsically linked to levels of livestock ownership is householdsize, expressed in terms of numbers of wives and children per husband. These data have beenreported fully elsewhere (Coast, 2000).

P1: GVM/GDX



main groups. First, household organization is outlined, forming the firstcomparative study of Maasai household size and arrangement. Second, anattempt is made to sketch contemporary living conditions of Maasai house-holds using housing type and drinking water source. Both variables arestrongly linked to outcomes of morbidity and mortality, and provide a proxyfor the context of Maasai living conditions. Third, Maasai livelihoods aredescribed using a combination of indicators: practice and type of cultiva-tion, practice of transhumance, and occupation. These data allow for botha description of contemporary Maasai economic activity, and for an assess-ment of the extent of diversification both within and between Kenya andTanzania. The free-answer response to occupation also allows for a detailedconsideration of two further specific aspects of livelihood: the participationof Maasai in tourism and the prevalence of murranism.5 Murranism is a tra-ditional occupation of Maasai men, and measuring it provides an importanttool with which to assess change within Maasai society. Finally, as a key com-ponent of current and future human capital, education levels are reported.Where appropriate, indicators are presented with reference to gender in or-der to highlight the different contemporary experiences of Maasai men andwomen. Before describing the results of this study, the next two sections de-scribe the broad-scale demographic context of Maasai and the methodologyused to collect the data.

MAASAI TOTAL POPULATION

Both Kenya and Tanzania conduct national censuses, the most recentbeing 19996 and 1988,7 respectively. However, in terms of estimating thetotal number of Maasai, only the Kenyan censuses collect ethnicity dataas part of the census schedule. An explicit question on ethnicity has notappeared in recent Tanzanian censuses, therefore it is not possible to estimateaccurately the number of Tanzanian Maasai.

Maasai are numerically the largest single group of pastoralists in Kenya.8

In Narok and Kajiado, the “Maasai districts,” the proportion of Maasai has

5A murran (warrior) describes a stage in a youth’s life when he has been circumcised, anda new age set produced. Over a period of time, all of the boys (who have usually reachedpuberty) are circumcised and incorporated into the newest age set. Circumcised young menare junior warriors, a traditional period associated with the establishment of manyattas, acamp to protect their neighborhood. For detailed descriptions see, for example, Spencer, 1988.

6To date, however, the 1999 Kenya census data are unavailable, therefore figures from the 1989census are used.

7The 1998 Tanzanian census was postponed, and has yet to be carried out.8Grandin states that “The Maasai are the second biggest group of pastoralists in Kenya, afterthe Somali” (1991, p. 21). The Kenyan census data refer only to Maasai whereas Somali arereported using a variety of subgroups including Gosha, Hawiyah, Ogaden, Ajuran, Gurreh,and Degodia.

P1: GVM/GDX


84 Coast

Table I. Maasai Totals, Kenyan Census Data (1962–1989)

Kenya Kajiado Narok

1962Maasai as % total administrative population 1.8% 78% 79%Total Maasai 154,079 78,887 86,472



declined substantially over the last 30 years, because of in-migration by non-Maasai. Indeed, by 1989, Maasai accounted for less than half of the districtpopulation in Narok (Table I).9

In Tanzania, the NCAA has carried out periodic “headcount” surveysof total population in the area. A common feature of all of these censusesis the poor reliability of the data, due to both internal and external factors.For example, Kijazi et al. state that the 1987 census was “conducted at a timewhen relationships between the NCAA and NCA residents were particu-larly strained . . .Previous talk of the feasibility of removing the Maasai fromthe NCA entirely are also likely to have discouraged NCA residents fromaccurately reporting total numbers in the NCA at this time” (1997, p. 174).The most recent NCAA census enumerated 51,621 individuals, of whom itis estimated that over 97% are Maasai (NCAA, 1999).

METHODOLOGY

A single round household survey (SRHS) was conducted betweenOctober 1997 and May 1998. Information was collected on 14,928 individualsat three main sites using a standardized questionnaire administered at thehousehold level by locally recruited and trained Maasai enumerators. Theabsence of complete and up-to-date sampling frames prevented systematicscientific sampling of the population. Study sites were chosen specifically toreflect the range of experience of rural Maasai (see Fig. 1).10

An enumeration unit must have some real, tangible meaning for theresearcher, enumerator, and respondent. The choice of enumeration unitis therefore a compromise between one that is ethnographically specificand one that conforms to some standardized schema. The standard United

9Although no data are yet available from the 1999 census, it is expected that Maasai will soonaccount for less than half of the Kajiado district population.

10The majority of the Maasai population in both Kenya and Tanzania is rural.

P1: GVM/GDX



Fig

.1.S

tudy

site

loca

tion

.

P1: GVM/GDX


86 Coast

Nations definition of a household as “One or more persons who make com-mon provision for food and other essentials for living” was inappropriate ina Maasai context. If the provision of food in a Maasai context is examined,then a complex web of normative food-sharing practices is revealed. Forexample, in a polygynous family each wife prepares and cooks food (such asmaize porridge) in her own house (enkaji), for consumption by herself andany dependants. Except in special circumstances such as sickness or prepa-ration for a ceremony, co-wives do not make common provision for cooking.There is no single word in Maa that corresponds precisely to “household.”More often, the term olmarei (family) is used, and it is clear from the contextthat it is the “household” that is meant.

Ideally, the primary sampling units (in this case the household) shouldbe randomly selected. However, a scientifically designed random sampleassumes the existence of a complete and correct sampling frame. Neither inKenya11 nor in Tanzania12 were such sampling frames available. Therefore,a geographic area was identified, and all sampling units within that area wereinterviewed, until the desired sample size was achieved. The following foursections describe the substantive results of the SRHS.

LIVING ARRANGEMENTS

The Maasai enkang may be described as a joint residential unit com-posed of several households (olmarei). Homewood and Rodgers describethe enkang as the “Maasai ideal,” allowing for cooperation over grazing andherding decisions. The rationale for membership of an enkang is not neces-sarily based on familial affiliation and the overall composition may changeperiodically.

Early studies observed enkang to be composed of 6–12 households(Jacobs, 1965), but there appears to be an increasing tendency toward

11The national sampling frame, derived from census returns, and maintained by the CentralBureau of Statistics was unavailable for this study (Central Bureau of Statistics, pers. comm.).An alternative sampling frame was investigated, that of the registers of the Group Ranches.However, these lists were drawn up at the time of the formation of the Group Ranches andare now obsolete. Further, they only ever contained the names of the registered members(men) rather than a complete listing of household members. The highly political nature ofcurrent access to Group Ranch land would have made the construction of more up-to-datelists extremely difficult.

12Initially in Tanzania I hoped to use the “10 cell” system introduced during ujamaa wherebyeach registered village is composed of 10 units, each with its own leader. Theoretically, each10-cell leader should provide an up-to-date list of village residents, permitting systematicsampling. However, when this approach was tested in Endulen (NCA) during the pilot study,it was found that the majority of lists was out of date and poorly maintained. Further, thenumber of households per 10-cell leader varied greatly due to a lack of candidates to be 10-cellleaders.

P1: GVM/GDX



Table II. Extant Research on EnkangSize Since 1950, by Country

Year(s) Tanzania Kenya

1950 6.2a

1960s 7–9b 4.0c 5.11970s 2.5d 1.7 2.71978 3.21980 5.3 2.5e

1983 1.81984 5–7 f

1985 1.31993 3–5aNjoka (1979), quoted in Grandin(1986).

bJacobs (1978), Tanzanian Maasailand.cArhem (1985), NCA.dNdagala (1982), Monduli.eGrandin (1986), Kajiado.f Potkanski (1993), Ngorongoro.

single-household enkang. This trend has been noted by several authors(Ndagala, 1982; Grandin, 1991; Potkanski, 1993). Potkanski suggests thatthe process toward single-household enkang is more common among wealth-ier families as a “result of ecological factors, since the number of animalsstaying together in one homestead should not be too many” (1993, p. 31),together with an increasing tendency toward the “individualization of so-cial life” (ibid), especially for wealthier families. Grandin (1991) and Rutten(1998) regard individual land ownership and greater levels of sedentariza-tion as crucial in facilitating the trend toward fewer households per enkangin Kenya. Table II summarizes all the existing information available on thenumber of households per enkang.

The influence of changing external influences on enkang size may beseen clearly in Tanzania. For example, ujamaa caused a decline in averageenkang size in the 1970s as people were moved to comply with the policyof permanent settlement. Homewood suggests that the later reversal of thisinitial decline implies that “families . . . split up to fulfil the imposed require-ments of ujamaa village membership, while retaining access and rights to abroad range of pasture resources, then reverted when the pressure relaxed”(1995, p. 342). The SRHS collected information on enkang composition, al-lowing for the first time a comparable cross-border survey on enkang size(Table III).

At the country level, Tanzanian enkangs are composed of more house-holds than in Kenya, with a mean size of 4.1 and 2.6 households per enkang,respectively. This overall pattern concurs with the extant data summarizedin Table II. At both Kenyan sites, single-household enkang are the most

P1: GVM/GDX


88 Coast

Table III. Number of Olmarei Per Enkang, by Major Study Site, SRDS Data

Mean Median Mode Minimum Maximum n

KenyaKoyaki G.R. 2.8 2.0 1.0 1.0 8.0 182Olkirmatian/Shompole G.R. 2.6 2.0 1.0 1.0 10.0 453All 2.6 2.0 1.0 1.0 10.0 635

TanzaniaEndulen–Esere 4.4 3.0 1.0 1.0 19.0 502Irkeepus 3.3 3.0 3.0 1.0 7.0 231Meshilli 4.1 4.0 4.0 1.0 11.0 186All 4.1 3.0 1.0 1.0 19.0 919

Note. G.R.: Group Ranch.

common. Extreme variations in enkang size are found at Endulen-Esere(Tanzania), where the modal enkang is a single-household unit, but therewas one 19-household enkang.13

The household in physical terms refers to the collection of houses abouta communal gate. The gate carries the name of the man and a separate gate isa symbol of his autonomy as a cattle owner and founder of a family. Grandindescribes the household as “the primary unit of production” (1991, p. 21).

Households in Tanzania are generally smaller than those in Kenya, al-though household sizes in the Olkirmatian/Shompole study site are verysimilar to those found in the Tanzanian study sites (Table IV). The largestenumerated household was composed of 88 people, at Endulen-Esere(NCA). It should be noted, however, that such large households are rela-tively rare, normally belonging to a very powerful individual such as a laibon.Similarly, while single-person households were recorded in both Kenya andTanzania, they are also very rare.14

LIVING CONDITIONS

The type of housing occupied by a household reflects three factors:household wealth, local polices restricting certain housing types, and thelevel of household transhumance. Maasai houses (enkaji) are traditionallymade from a mixture of dung and mud, smeared on a wooden framework.Every married woman eventually has her own house, and it is her respon-sibility to build and maintain it for herself, her children, any dependants,and (periodically) her husband. At the level of the house, Maasai women

13Although it should be noted that this was the enkang belonging to a laibon (a traditionalhealer).

14A total of three single-person households were enumerated, representing 0.2% of all house-holds included in the SRHS (n = 1545).

P1: GVM/GDX



Table IV. Household Size, by Major Study Sites, SRDS Data

Mean Median Mode Minimum Maximum n

KenyaKoyaki G.R. 12.9 10.0 6.0 2 63 182Olkirmatian/Shompole G.R. 9.2 7.0 6.0 1 80 453All 10.3 8.0 6.0 1 80 634

TanzaniaEndulen–Esere 8.7 7.0 5.0 2 88 502Irkeepus 10.6 8.0 6.0 2 56 231Meshilli 8.2 7.0 3.0 1 33 186All 9.1 7.0 5.0 1 88 920


possess relatively high levels of autonomy in terms of domestic affairs andthe family economy (Talle, 1987).

Recently, however, there have been changes in traditional Maasaihouses. The “modernization” of housing has become common (Ndagala,1982), taking the form of rectangular structures with corrugated iron orthatched roofs (Kipuri, 1989). Changing house styles also represent a shiftaway from houses being associated solely with women. Iron roofs requiremoney to buy the building materials, so it is more likely that a man is able tomake the purchase, and the need to measure and cut materials means thatspecialists in house building are brought in from outside the family unit. Itis becoming increasingly common for a man to build himself a house as astatus symbol and not necessarily for the use of his family.15 Almost half ofenumerated households in Koyaki possessed at least one building with aniron roof, compared with only 8% of households in Olkirmatian/Shompole.Levels of ownership of iron-roofed buildings are very low (3.6%) at the NCAstudy sites16 (Table V).

The degree of permanence of building structures has important im-plications for a traditionally transhumant society such as the Maasai. Aniron-roofed house may be both a cause and an effect of increasing levels ofsedentarization. As a cause, permanent structures will require year-roundresidence by at least some members of the household. The increased levels ofpermanent buildings, particularly in Koyaki Group Ranch, may represent aneffect of changing livelihoods away from traditional transhumant pastoral-ism toward formal employment and/or cultivation. Exogenous influences

15Although Kipuri suggests, “in cases where a man builds a western style house but has severalwives, the favourite one normally is asked to move into the western house” (1989, p. 306).

16It should be noted, however, that the building of “permanent” accommodation is subject tovery strict rules in NCAA, which has its own set of planning and building regulations. Therehas been a change toward thatched roofing materials in NCAA, away from the traditionalmud/dung plaster. No data were collected on the prevalence of roofing thatch.

P1: GVM/GDX


90 Coast

Table V. Households Owning At Least One Building With an Iron(Mabati ) Roof

Percentage ofhouseholds n

KenyaKoyaki–Lemek G.R. 47.8 182Olkirmatian/Shompole G.R. 8.0 448Total 19.6 630

TanzaniaEndulen–Esere 6.0 500Irkeepus 0.4 229Meshilli 1.1 186Total 3.6 915


such as NGOs17 have also played a part in the transformation of Maasaienkaji in selected locations. The implications of poorly ventilated enkaji formorbidity have been noted (AMREF, 1999).

The source of drinking water for a household is an important determi-nant of health, and seasonal variations in availability of drinking water can re-flect an insecure drinking water source. Fetching drinking water is essentiallya female task, therefore the implications of an insecure drinking water sourcefor a woman’s work budget are particularly important (Grandin, 1991).

Surface water sources provide most of the water for all the study sites(Table VI). Irkeepus has seasonal access to spring water, although recentlythere have been problems with the security of this water source because ofofftake by a neighboring tourist lodge. Olkirmatian and Shompole GroupRanches have the most diversified drinking water sources, because of thepresence of the Shompole swamp combined with recent water infrastructuredevelopments for the export vegetable crops.

LIVELIHOOD

Early accounts of the Maasai by travellers and colonial administra-tors tended to portray an image of “pure” pastoralism, reliant on live-stock products for subsistence (e.g., Hollis, 1905; Merker, 1910; Leakey,1930; Fosbrooke, 1948). Such accounts have been largely disregarded andreferences to cultivation by Maasai have a long history (Berntsen, 1979).Spear notes that even before the eighteenth century “many Maasai prac-tised a mixed agropastoral economy” (1993, p. 131). Indeed, the abandon-ment of livestock herding in favor of settled cultivation by the Maa-speaking

17For example, the Maasai Housing Project run by Intermediate Technology DevelopmentGroup (ITDG).

P1: GVM/GDX



Table VI. Percentage Distribution of Drinking Water Source, By Study Site and Season

Kenya Tanzania

Olkirmatian/ Endulen–Koyaki G.R. Shompole G.R. Total Esere Irkeepus Meshilli Total

Wet seasonDam 1.1 13.8 10.0 0 0 45.7 9.6Well 0.0 12.9 9.2 0 0 0 0.0River 66.5 38.8 46.9 44.8 93.0 15.6 50.9Pond 1.1 15.2 11.1 0 0 0.5 0.2Spring 30.2 0 8.7 50.8 7.0 38.2 36.9Tap 1.1 19.2 14.0 4.4 0 0 2.4

Dry seasonDam 1.1 0.0 0.3 0 0 34.4 7.1Well 0 0.2 0.2 7.4 0 0.5 4.1River 37.9 71.2 61.5 88.2 82.1 62.4 81.2Pond 16.5 2.2 6.4 0 0 0 0Spring 40.1 0.0 11.6 0 17.9 2.7 5.1Tap 4.4 26.3 20.0 4.4 0 0 2.4


WaArusha (who retain many elements of “Maasai” social organization) isperhaps the most extreme example of this (Spear and Waller, 1993). The trad-ing of livestock for grain has been described extensively (Dahl and Hjort,1976; Swift, 1986), together with detailed studies of Maasai diet that demon-strate the role played by nonlivestock products (Arhem, 1985; Nestel, 1986;Bekure et al., 1991).

It is clear that subsistence-grown and bought grain play a (major) rolein contemporary Maasai diets, as milk yields decline with the onset of thedry season, crops become ready for harvesting. However, the developmentof cash crop cultivation by Maasai has received less attention in the litera-ture. Throughout all of the study sites, cash and subsistence cultivation werepresent to varying degrees (Table VII). There is great variation (determinedboth by policy and agroecological potential) in the type of cultivation thatpredominates in each location.

Cultivation of cash crops predominates in two locations: Olkirmatian(Kenya) and Irkeepus (NCA). Elsewhere, cultivation is far more episodicand sporadic, reflecting interannual weather variations. For example, inOlkirmatian Group Ranch there has been rapid (last decade) developmentof cash-crop cultivation specifically for the export market. The presence ofperennial water sources allowing for rain-fed irrigation at the foot of theNguramen escarpment combined with good road access to Nairobi help toexplain this development. A similar example is Irkeepus (NCA) where eco-logical conditions (cool uplands with >1,000 mm precipitation per annum)combined with demand from tourist facilities have driven the developmentof small-scale vegetable cultivation.

P1: GVM/GDX


92 Coast

Table VII. Major Cultivation Types, by Study Site

Cultivation types

KenyaKoyaki G.R. Subsistence grain cultivation + some larger-scale wheat

cultivationOlkirmatian G.R. Mainly commercial cultivation for export market+ small amount

of subsistence grain cultivation + perennial rain-fed irrigationShompole G.R. Subsistence gain cultivation + some irrigation from Shompole

swampTanzania

Meshilli Subsistence grain and bean cultivationIrkeepus Vegetables for sale to tourist lodgeEndulen–Esere Subsistence grain cultivation


Table VIII shows the percentage distribution of cultivation at the house-hold level at the time of the survey and 10 years previously. Taken at facevalue, the results from the study sites in NCA would imply a massive andvery rapid uptake of cultivation. However, the results are an artifact of thefact that 10 years prior to the SRHS, cultivation was illegal within NCA.18

What the SRHS shows is a high level of involvement of Maasai house-holds in some form of cultivation, both subsistence and commercial. Over88% of the NCA households and 45% of the Kenyan households reportedcurrent cultivation. Maasai cultivate for a wide range of reasons, from sub-sistence cultivation of beans designed to complement a pastoral subsistencestrategy to intensive rain-fed irrigation of export cash crops with a sophisti-cated system of marketing and transportation. Cultivation, to varying de-grees, is associated with a more sedentary lifestyle as continual labor isrequired to sow, weed, and harvest crops. For a traditionally transhumantpastoralist population such as the Maasai, the uptake of cultivation has im-plications for population mobility, providing a further measure of changinglivelihoods.

The importance of transhumance for pastoralism as part of a strategy tocope with the high degree of interannual variability in the savanna ecosys-tem has been well documented (e.g., Behnke and Scoones, 1993). Severalauthors note increasing levels of sedentarization by the Maasai, includingGrandin (1991), Fratkin (1994), and Rutten (1998). Reasons for increasedsedentarization include individual land tenure (and the erection of fences),cultivation, and increased use of education services. All individuals agedover 6 years were asked whether they had moved with their herds for more

18Although it should be noted that cultivation continued to be practiced. However, the highlypoliticized nature of the cultivation ban meant that individuals tended to deny any previouscultivation.

P1: GVM/GDX



Table VIII. Percentage of Households Reporting Cultivation byTime Period and Study Site

Cultivated Currently10 years ago cultivate

KenyaKoyaki G.R. 15.9 31.3Olkirmatian/Shompole G.R. 20.1 51.8Total 18.9 45.9

TanzaniaEndulen–Esere 1.1 98.2Irkeepus 5.7 76.4Meshilli 0.0 76.3Total 1.9 88.2


than 1 week in the preceding 12 months. While it is recognized that the SRHSwas cross-sectional in design, and is not necessarily a reflection of the inter-annual variation in transhumance, of relevance here is the relative practiceof transhumance between the different study sites. Because pastoralism isa household endeavor, it is pertinent to examine transhumance both at theindividual and the household level, particularly when considering the im-plications of cultivation for the practice of transhumance. A household thatpractices no cultivation is much more likely to move as an entire householdthan is a cultivating household, where some family members will have toremain near the fields. There is extreme variation in transhumant migrationboth among and within the study sites (Table IX). For example, over 90%of households in Irkeepus reported no household members practicing tran-shumant migration, compared with Olkirmatian/Shompole Group Rancheswhere over 33% of households migrate as an entire household.

Table IX. Household-Level Transhumance, by Study Site

Transhumance

Partial WholeNone householda household

KenyaKoyaki–Lemek G.R. 69.2 30.2 0.5Olkirmatian/Shompole G.R. 42.4 22.5 35.0Total 50.2 24.8 25.0

TanzaniaEndulen–Esere 4.8 68.0 27.2Irkeepus 92.6 6.6 0.9Meshilli 2.2 96.2 1.6Total 26.2 58.3 15.5

Note. G.R.: Group Ranch.aAt least one member of the household migrated.

P1: GVM/GDX


94 Coast

Fig. 2. Percent distribution of occupation, all individuals aged over 15 years, by village andcountry.

The influence of the ecological context must be taken into account. Ofall the study sites, Olkirmatian/Shompole is the driest and therefore moreexposed to interannual rainfall variability. However, the influence of chang-ing land tenure in Kenya must also be noted. Olkirmatian/Shompole GroupRanches are not in the process of subdivision of land, and communal grazingcommittees continue to operate. In comparison, the distribution of individualtitle deeds to land is well under way in Koyaki Group Ranch. That individualownership of land is incompatible with transhumance has long been recog-nized. Within the Tanzanian study sites, where the variable of land tenure isessentially “controlled,” variations in the occurence of transhumance reflectecological conditions very clearly. In upland Irkeepus there is virtually notranshumance compared with the much drier lowland Meshilli, where sea-sonal movements toward the Gol Mountains have been noted (Thompson,1997).

Quantitative data on cultivation and transhumance provide the contextwithin which individual occupations might be placed. Questions on occu-pation were framed very loosely and referred to the amount of time anindividual spends on a particular task, rather than any consideration of theamount of income derived from that activity. The data contained in Fig. 2refer only to individuals aged 15 years and above.19

19And does not include students in the total.

P1: GVM/GDX



The distribution of occupations is highly variable between locations,reflecting both economic opportunities and ecological influences. In Kenyathe opportunities for diversification away from “traditional” pastoralist ac-tivities are greater than in Tanzania. However, it should be noted that for themajority of individuals, pastoralism still plays some part in their livelihood,albeit in combination with some other activity. In Tanzania, only at Endulen-Esere are there opportunities for diversification of occupation, reflecting thepresence of a hospital, trading center, police station, and veterinary center.

The influence of cash crop opportunities in Olkirmatian/ShompoleGroup Ranches is clearly highlighted, with 8.3% of individuals reportingcultivation of cash crops as their only occupation. It is likely, of course, thatthese individuals still remain integrated to some degree within the pastoraleconomy as they are often younger (educated) sons of traditional pastoralistfamilies.

Women tend to have far less diversified income sources than men. Oneof the main ways in which it is (traditionally) acceptable for a Maasai womanto earn cash income is to sell milk that is surplus to household require-ments (Table X), although the ability to sell milk varies seasonally. Themaking and selling of beadwork by women reflects the presence of an op-portunity; Koyaki Group Ranch being so close to the Maasai Mara has aready market for beadwork. The majority of women make the beadwork butonly a few women actually sell the work directly to tourists, either at parkgates or at cultural bomas. Most beadwork is sold by middlemen to touristshops and traders, although no data on this were collected (see, e.g., Kipuri,1998).

The only other significant source of income for the women in the studywas shopkeeping. This often involved very small stores attached to the homewhere a very limited stock of essentials were sold (fat, flour, drugs, sweets).However, it is likely that most of these stores are in fact owned and financedby a husband or male relative and the women simply attend to the stores.

A second gendered element of occupation is that of murranism. Manyethnographies refer to the period of being a murran (warrior) as if it were afull-time occupation for Maasai men aged approximately 15–30. However,this is possibly more an overemphasis of the symbolic role of murran than a

Table X. Percentage of Women Deriving Some Income FromMilk and Beadwork Sales

Milk selling Beadwork

Koyaki G.R. 0.2 20.5Olkirmatian/Shompole G.R. 4.4 0.3NCA 20.8 0.9


P1: GVM/GDX


96 Coast

Table XI. Numbers of Individualsa Who Described EitherTheir Main or Secondary Occupations as Murran, by Major

Study Site

Total Murran %

KenyaNarok 285 0 0Kajiado 539 84 15.6All 824 84 10.2

Tanzania 982 135 13.7aMen aged 15–30 years.

true reflection of contemporary Maasai lifestyle. Because it was possible fora man to reply “murran” in the SRHS, it is possible to explore contemporarylevels of active participation in this traditional occupation (Table XI).

While still very important symbolically and socially, the “active” role ofbeing a murran (e.g., defending property) is no longer important on a day-to-day basis.20 Although there are no comparable historical data, I suggestthat if similar data had been collected two or three decades ago, many moremen would have replied that their primary or secondary occupation was“murran.” Indeed, not a single individual reported themselves as being amurran at the Koyaki Group Ranch study site, the location with the highestlevels of economic diversification and formal employment opportunities.21

Unsurprisingly, murran with their distinctive hair and jewellery, areone of the main features associated with Maasai by the large numbers ofinternational tourists that visit Maasailand. Spear suggests that the inter-national recognition of Maasai is due to “countless coffee-table books andtourists’ snapshots” (1993, p. 1). The importance of international tourismfor the Kenyan and Tanzanian economies is well established. For example,the revenue from international tourism is estimated to contribute 19% ofKenya’s Gross Domestic Product (World Bank, 2001). Two of the study sitesare adjacent to major international tourist destinations: Masai Mara GameReserve (Koyaki Group Ranch) and Ngorongoro Crater (NCA sites). Issuesof marginalization of the majority of Maasai from the income-generatingopportunities provided by tourism are increasingly being raised (Kipuri,

20Although murran may still be called upon at times of need, and in Kenya are increasinglybeing associated with local politicians (Hodgson, 1999). Livestock raiding continues to be anillicit activity in some locations, particularly in western NCAA. Some aspects of murranismare still important, including the long-distance, possibly cross-border herding of livestock.

21The data are cross-sectional, however, and it should be noted that many individuals may havereported themselves as murran at a different time in the year. This is because periodic large-scale age-set murran events do take place and form an important part of the social calendar. Ifone of these events had recently taken place, then higher numbers of murran would probablyhave been reported. These results do not imply the diminution of the significance of being amurran and being involved in an age-set succession process.

P1: GVM/GDX



Table XII. Tourism-Related Employment Opportunities, by Sex

Men and women Cultivate and sell vegetables to lodges, lodge entertainment (singingand dancinga), work in cultural manyatta

Men Game ranger/warden, campsite/lodge guard, lodge room cleaner,keep bees and sell honey to lodges, trader in beadwork, touristguide, tour driver

Women Make beadwork, sell beadwork directly to touristsaIt should be noted that many of the people who perform the singing and dancing are non-Maasai.

1998; Akama, 1999). It must be made clear that the present discussion ex-cludes measurement of actual income derived from tourism. Further, be-cause involvement in tourism here is limited to occupations, indirect sourcesof tourist-related income cannot be addressed.22 However, very few peo-ple derive all their income from tourism alone. In NCA, 0.2% of all in-dividuals aged over 15 were employed full-time in the tourism industry.In Koyaki, just 1.3% of the sample population were full-time tourism em-ployees. Of course, these figures are no more than the most cursory de-scription of employment opportunities within the tourist sector, but they doreinforce the view that the majority of Maasai who live near tourist attrac-tions do not benefit directly.23 Particularly in Koyaki, it was interesting tonote that individuals with full-time jobs in tourism tended to be clusteredin the same households. There are marked differences in the opportunitiesafforded by the tourism industry for men and women, as summarized below(Table XII).

Indirect opportunities from tourism must also be noted. For example, inIrkeepus (NCA), 27.3% of adults report that they grew vegetables (cabbages,carrots, and potatoes) in order to sell them to the nearby tourist lodge.No information can be presented here on the income derived from thesetransactions, but discussion with key informants suggested that prices offeredby the lodge are very low. Other indirect sources of tourism-related incomeinclude the making and selling of beadwork (both to traders and direct totourists) and the sale of honey to tourist lodges.24 “Cultural bomas” operatein Talek, Irkeepus, and Meshilli but the advantages tend to accrue to only afew selected individuals. Indeed, much of the entrance fee for cultural bomas

22Substantial returns from tourism for a few individuals have been noted, particularly from theleasing of land to tourist developments in Narok District (personal observation, Thompson,pers. comm.).

23Although it should be noted that the financial opportunities available to local Maasai in thestudy areas are greater than those available to communities living near other protected areasin Kenya and Tanzania, mainly because of the size of the tourist revenue generated by thesetwo protected areas.

24In Irkeepus and Meshilli (NCA) in response to NGO equipment and training provision.

P1: GVM/GDX


98 Coast

Fig. 3. Percentage distribution of individuals with completed primary school education by fiveyear age group, sex and country.

is often diverted at source by tour guides, and only a small proportion isretained by the Maasai (Kipuri, 1998).

EDUCATION

Participation in formal education is a function both of supply and de-mand. The problems of providing schooling in remote rural areas with poorinfrastructure are well known, particularly for nonsedentary populations(Swift, 1990).25 Several studies of Maasai participation in formal educationhave been carried out (King, 1972; Gorham, 1980; Holland, 1996), all report-ing low levels of school attendance by eligible children. When combined withthe traditional antipathy of the Maasai to sending their children to school,26

the low levels of education reported for older Maasai in the survey are un-surprising (Fig. 3). The sex bias in completed primary school education ismarked, in both countries, with far fewer women than men having attendedschool.

25The problems include lack of demonstrable benefits of education for pastoralism, labor re-quirements of herding, for transhumant settlement patterns, relatively high (and increasing)school fees, lack of education infrastructure, low population densities, and the reluctance ofteachers to live in relatively isolated areas (after Gorham, 1978).

26Described by Galaty as “With education, Maasai often feel they have given away their chil-dren” (1992, p. 37).

P1: GVM/GDX



There is evidence of an improvement in access to education over time.However, it is interesting to note the overall decline in the levels of educa-tion for Tanzanian Maasai men, from 11.0% of men aged 50–54 to 4.0% ofmen aged 25–29 years. Perhaps most informative are current levels of en-rolment for Maasai children, relative to national rural levels. Of all childrenaged 7–12 years, 32% of Kenyan Maasai were currently attending school,relative to the national rural average of 65%. The corresponding figures forTanzania are even more dismal, with only 9% of 7–12 years old Maasai cur-rently attending school, relative to the national rural average of 47% (UNDP,1999). That Maasai are marginalized in terms of access to education—oneof the key determinants of future human capital—is underlined in the datapresented here.

DISCUSSION

What do these large-scale data tell us about contemporary Maasai liveli-hoods and living conditions? Maasai have persisted as livestock owners, withover 98% of all households owning livestock.27 In Kenya and Tanzania, 90.1and 98.8%, respectively, of adults describe themselves either as pastoral-ists or agropastoralists. That cultivation is playing a large role in householdlivelihood strategies is also evident, although the variety in types of andmotivations for cultivation must be highlighted.

In Kenya the uptake of cash crop cultivation has been extremely rapidwhere conditions allow. This is particularly so in the Shompole/OlkirmatianGroup Ranches, where ecological conditions permit rainfed irrigation ofcash crops and over half of all households cultivate for the export market. InTanzania, the use of cultivation for cash crops is also evident, particularly forsale to the tourist lodges and NCAA staff, although on a smaller scale and formuch lower returns than in Kenya. The extensive exploitation of cash cropcultivation by Maasai in both Kenya and Tanzania demonstrates widespreaddiversification of the subsistence economy, wherever conditions allow. Whileprevious work on Maasai cultivation has tended to focus on subsistencepurposes, the data from the current study show widespread cultivation forfinancial return in both countries. Although clearly limited by agroecologicalconditions and the existence of a market, the uptake of cash crop cultivationby Kenyan Maasai households has been rapid. It is less easy to determinethe rate of cultivation uptake in the Tanzanian study sites, due to the ban oncultivation in the NCA, which was only lifted in 1992 following a directive

27The absence of livestock data make it impossible, though, to comment on the scale of thatlivestock ownership.

P1: GVM/GDX


100 Coast

from the Office of the Prime Minister. However, cultivation within the NCAis widespread and has been the focus of much controversy in this multiple-land-use area (McCabe et al., 1997).

By combining the individual and household data on cultivation, it canbe demonstrated that at the country level there are significant differences inthe way in which labor for cultivation is organized within the household. InKenya only 15.7% of all adults include cultivation in the description of theiroccupation, yet 45.9% of all households practice some cultivation, suggestingthat it tends to be specific individuals in each household who do the cultiva-tion. In Tanzania, on the other hand, 72.1% of all adults practice some formof cultivation, distributed across 88.2% of all households. This country-leveldivergence in the organization of cultivation is possibly linked to differentmotivations for cultivation. In Kenya, the cultivation of cash crops at a plotsome distance from the enkang necessitates that a portion of the house-hold (perhaps an elder son or a wife) is responsible for the cultivation, sale,and transportation of the produce. In Tanzania cultivation is dominated bysmall-scale cultivation, either next to the enkang or a short distance away.Here, a larger proportion of the household will be involved in the day-to-daycultivation of the plot, and there is less demand for specialization within thehousehold.

Cultivation is an important subsistence strategy for the Maasai. In thewake of increasing food supply insecurity, triggered by the combination of arapidly growing human population and a static or declining livestock popu-lation, trends toward the cultivation of subsistence crops (particularly maizeand beans) are to be expected. This study underlines the importance ofsubsistence cultivation for both Kenyan and Tanzanian Maasai, support-ing the findings of other researchers (Arhem et al., 1981; McCormick andElmore-Meegan, 1992).

The complex interrelations among different aspects of Maasai liveli-hood and lifestyle become apparent when an attempt is made to disentanglethe different forces at work in changing living arrangements, building ma-terials, and frequency of transhumance. For example, the organization ofhouseholds and enkang diverge noticeably between Kenya and Tanzania.Put simply, Kenyan Maasai are more likely to live in larger single house-holds whereas Tanzania Maasai are more likely to live in smaller multiplehouseholds. This difference cannot be explained in terms of different ratesof polygynous household formation, as there is no significant difference ineither the prevalence28 or intensity29 of polygyny between the Kenyan and

28Prevalence of polygyny refers to the proportion of men in polygynous marriages, and was46% for both Kenya and Tanzania in the SRHS.

29Intensity of polygyny refers to the average number of wives per polygynist, and was 2.72 forKenyan Maasai and 2.84 for Tanzanian Maasai in the SRHS.

P1: GVM/GDX



Tanzanian Study sites (Coast, 2000). The dominance of single householdenkang in Kenya can be linked to other changes in Maasai lifestyle, includ-ing the high levels of iron-roofed permanent housing and relatively low levelsof whole household migration.

Closely bound up with relatively lower levels of transhumance in Kenyaare wider sociopolitical developments, such as the increasing individualiza-tion of land tenure resulting in the construction of fences that operate asphysical barriers to livestock movement. When all of these elements arecombined (iron-roofed house plus cultivation plus restricted livestock move-ment), seen perhaps most clearly at the Koyaki study site, it is unsurprisingthat livelihood diversification is occurring. Tanzanian Maasai living in NCA,by way of contrast, do not have the same restriction of livestock movementand levels of iron-roofed housing ownership are still low.30 The result is thatpartial household migration dominates. Any discussion of transhumance as asurvival strategy ignores the ecological context at its peril. In those study sitessituated in drier locations, particularly Olkirmatian/Shompole and Meshilli,migration of either part or all the household remains an important tactic.

The different spheres of male and female Maasai life have been con-sidered extensively, particularly in the work of Talle (1987, 1999), Llewelyn-Davies (1978), and von Mitzlaff (1988). The data from the SRHS permit sometentative quantitative assessment of the differences in occupation and edu-cation (themselves strongly linked) between men and women. That Maasaiwomen are far less integrated than men within the “nontraditional” eco-nomic system and have lower levels of educational attainment isunsurprising—the same pattern is found throughout sub-Saharan Africa andmost of the developing world. It is perhaps in the tourist industry that womenare most excluded relative to men. The opportunities for direct employmentare extremely limited for all Maasai, and the narrow range of jobs avail-able for women reduces their potential to engage in this arena even further.Low female participation in formal education further reduces the chancesfor employment outside of the household, as an ability to at least speakKiSwahili and preferably also English are prerequisites for such jobs. Thegreatest irony is that it is Maasai women’s handicrafts and jewellery thatare most associated with the Maasai by the international tourists that visitMaasailand, but few Maasai women benefit directly from this trade.

One drawback to a cross-sectional study such as its inability to cap-ture information on those individuals who have moved out of the traditionalrural Maasai system. Anecdotal evidence suggests that in recent years op-portunities for Maasai men (especially those with an education) outsidea “traditional” Maasai lifestyle have become increasingly accessible and

30See footnote 16.

P1: GVM/GDX


102 Coast

attractive. Given these expanding opportunities, it is possible that such menmove away permanently and then marry non-Maasai women. Spear, for ex-ample, suggests, “education and work elsewhere supplants the socializationof the murran” (1993, p. 14). The only circumstantial data to support thesuggestion that there is outmigration of Maasai males to urban areas is thatprovided by the most recent Kenyan Census.31 The sex ratio for Maasailiving in Nairobi has been very high for the last two censuses, at 270 and247 males per 100 females in 1979 and 1989, respectively. This implies sex-selective migration of Maasai men relative to Maasai women, although themovement of Maasai women due to destitution should not be ignored (Talle,1999).

That Maasai are becoming increasingly exposed to processes ofmarginalization and destitution has been noted by several authors (Arhem,1985; Hillman, 1994; Homewood, 1995; Talle, 1999). The very low levels ofprimary school enrolment relative to non-Maasai rural populations are alsocause for concern. In both Kenya and Tanzania, the diversification of in-dividual and household livelihood strategies will undoubtedly favor thosepeople with education, and not least the ability to speak the lingua franca,KiSwahili. Only three households in the current survey were recorded asdoing any available work in return for food, and had no livestock of theirown.32 Given the decreasing ability of Maasai and pastoralists in general todeal with subsistence crises, such cases will become less rare in the future.

REFERENCES

Akama, J. S. (1999). Marginalization of the Maasai in Kenya. Annals of Tourism Research 26(3):716–718.

AMREF (1999). An interim narrative report of an EU co-financed project: Community basedhealth care for nomads: Expansion of the Entasopia project in Kenya. Nairobi, AMREF.

Arhem, K., Homewood, K., and Rodgers, A. (1981). A pastoral food system: The NgorongoroMaasai in Tanzania. BRALUP Research Paper 70.

Arhem, K. (1985). The Maasai and the state: The impact of rural development policies on apastoral people in Tanzania, IWGIA Document.

Asiema, J. K., and Situma, F. D. P. (1994). Indigenous people and the environment: The case ofthe pastrol Maasai of Kenya. Colorado Journal of International Environmental Law 5(1):149–171.

Barkan, J. D. (ed.). (1994). Beyond Capitalism vs. Socialism in Kenya and Tanzania Rienner,Boulder, CO.

Behnke, R. H., and Scoones, I. (eds.). (1993). Range Ecology at Disequilibrium, Overseas De-velopment Institute, London.

31Comparable data for Tanzania are unavailable.32These families were referred to as Kibarua, a KiSwahili term used by Maasai to describe

casual work in return for food.

P1: GVM/GDX



Bekure, S., Leeuw, P. N. D., Grandin, B. E., and Neate, P. J. H. (1991). Maasai herding: An analysisof the livestock production system of Maasai pastoralists in eastern Kajiado District, Kenya,International Livestock Centre for Africa, Addis Ababa.

Berntsen, J. L. (1979). Economic variations among Maa-speaking people. Hadith: Ecology andHistory in East Africa 7.

Bonte, P., and Galaty, J. G. (1992). Herders, Warriors and Traders: Pastoralism in Africa, West-view Press, Boulder.

Coast, E. E. (2000). Maasai Demography, Unpublished PhD Thesis, University of London.Dahl, G., and Hjort, A. (1976). Having Herds: Pastoral Herd Growth and Household Economy,

University of Stockholm, Stockholm.Evangelou, P. (1985). Livestock Development in Kenya’s Maasailand: Pastoralist Transition to

a Market Economy, Westview Press, Boulder.Fosbrooke, H. A. (1948). An administrative survey of the Masai social system. Tanganyika

Notes and Records 26: 1–50.Fratkin, E. (1994). Problems of Pastoral Land Tenure in Kenya: Demographic, Economic and

Political Processes Among Maasai, Samburu, Boran, and Rendille, 1950–1990, AfricanStudies Center, Boston University, Boston.

Galaty, J. G. (1981). Land and livestock and Kenyan Maasai. Journal of Asian and AfricanStudies 16: 68–88.

Galaty, J. G. (1992). The land is yours: Social and economic factors in the privatisation, sub-division and sale of Maasai ranches. Nomadic Peoples 30: 26–40.

Gorham, A. B. (1978). The provision of education in pastoral areas. Pastoral Network Paper6B.

Gorham, A. B. (1980). Education and social change in a pastoral society: Government initia-tives and local responses to primary school provision in Kenyan Maasailand. Studies inComparative and International Education 3.

Grandin, B. (1986). Land tenure, sub-division and residential change on a Maasai Group Ranch.Bulletin of the Institute for Development Anthropology 4(2): 9–13.

Grandin, B. (1988). Wealth and pastoral dairy production: A case study from Maasailand.Human Ecology 16(1): 1–21.

Grandin, B. (1991). In Bekure, S., de Leeuw, P. N., Grandin, B. E., and Neate, P. J. (eds.),Maasai Herding: An Investigation of Pastoral Production on Group Ranches in Kenya,International Livestock Centre for Africa, Nairobi, Kenya.

Hedlund, H. (1980). Contradictions in the peripheralisation of a pastoral society: The Maasai.Review of African Political Economy 16: 15–34.

Helmut, R. (1995). Nachocho – an example of the change of the living sphere of the Maasai ofNorthern Tanzania. GeoJournal 36(1): 49–55.

Hill, A. G., and Randall, S. C. (1985). Issues in the study of Sahelian pastoralists and agro-pastoralists. In Hill, A. G. (ed.), Population, Health and Nutrition in the Sahel: Issues in theWelfare of Selected West African Communities, KPI, London.

Hillman, E. (1994). The pauperization of the Maasai in Kenya. Africa Today 4: 57–65.Hodgson, D. L. (1999). “Once intrepid warriors”: Modernity and the production of Maasai

masculinities. Ethnology 38(2): 121–150.Holland, K. (1987). Land, livestock and people: New demographic considerations for Kajiado

Maasai. East African Pastoral Systems Project Discussion Paper 5: 1–40.Holland, K. (1996). The Maasai on the Horns of a Dilemma: Development and Education,

Gideon S Were, Nairobi.Hollis, A. C. (1905). The Maasai, Clarendon Press, Oxford.Homewood, K., and Rodgers, A. W. (1991). Maasailand Ecology, Cambridge University Press,

Cambridge.Homewood, K. (1995). Development, demarcation and ecological outcomes in Maasailand.

Africa 65(3): 331–350.Hutchinson, J., and Smith, A. D. (1996). Ethnicity, Oxford University Press, Oxford.Ibrahim, B., and Ibrahim, F. (1995). Pastoralists in transition: A case study from Lengijape,

Maasai Steppe. GeoJournal 36(1): 27–48.

P1: GVM/GDX


104 Coast

Jacobs, A. H. (1965). The Traditional Political Organisation of the Pastoral Maasai, PhD Thesis,University of London.

Jacobs, A. (1978). A final report of development in Tanzania Maasailand: The perspective over20 years 1957–1977, USAID, Tanzania.

Jacobs, A. H. (1979). Maasai inter-tribal relations: Belligerent herdsmen or peaceable pastoral-ists? Senri Ethnological Studies 3: 33–49.

Kijazi, A., Mkumbo, S., and Thompson, D. M. (1997). Human and livestock population trends.In Thompson, D. M. (ed.), Multiple Land-Use: The Experience of the Ngorongoro Conser-vation Area Tanzania, IUCN, Gland, Switzerland.

King, K. (1972). Development and education in the Narok District of Kenya. African Affairs71(285): 385–407.

Kipuri, N. N. O. (1989). Maasai Women in Transition: Class and Gender in the Transformationof a Pastoral Society, PhD Thesis, Temple University.

Kupuri, N. N. O. (1998). Wildlife tourism and its impact on indigenous women in east Africa.In Vinding, D. (ed.), Indigenous Women: The Right to a Voice, International Work Groupfor Indigenous Affairs Copenhagen, Vol. 88, pp. 171–182.

Leakey, L. S. B. (1930). Some notes on the Masai of Kenya Colony. Royal AnthropologicalInstitute of Great Britain and Ireland 60: 185–209.

Llewelyn-Davies, M. (1978). Two contexts of security among pastoral Maasai women. In Caplan,J., and Bujra, J. M. (eds.), Women United, Women Divided: Cross-Cultural Perspectives onfemale solidarity Tavistock: London.

McCabe, T., Schofield, E. C., and Pedersen, G. N. (1997). Food security and nutritional sta-tus. In Thompson, D. M. (ed.), Multiple Land-Use: The Experience of the NgorongoroConservation Area Tanzania, IUCN, Gland, Switzerland.

McCormick, J., and Elmore-Meegan, M. (1992). Maasai diet. The Lancet 340.Merker, M. (1910). Die Masai: Ethnographische Monographie eines ostafrikanischen Semiten-

volkes. Translation from Catholic Mission, Ngong, Kenya.NCAA (1999). 1998 Aerial Boma Count, 1999 People and Livestock Census, and Hu-

man Population Trend Between 1954 and 1999 in the NCAA, NCAA, Ngorongoro,Tanzania.

Ndagala, D. K. (1982). Operation Imparnati: The sedentarisation of the pastoral Maasai inTanzania. Nomadic Peoples 10: 28–39.

Nestel, P. (1986). Nutrition of Maasai Women and Children in Relation to Subsistence FoodProduction, PhD Thesis, University of London.

Njoka, T. J. (1979). Ecological and Socio-Cultural Trends of Kaputiei Group Ranches in Kenya,PhD Thesis, University of California, Berkeley, pp. 318.

Potkanski, T. (1993). Pastoral Economy, Property Rights and Traditional Mutual Assis-tance Mechanisms Among the Ngorongoro and Salei Maasai of Tanzania, IIED,London.

Rutten, M. M. E. M. (1992). Selling wealth to buy poverty: The process of individualisation ofland ownership among the Maasai pastoralists of Kajiado District, Kenya, 1890–1990. InNimegien Studies in Development and Cultural Change, Nimegien, Breitenbach, Vol. 10,pp. 510.

Rutten, M. M. E. M. (1998). Land tenure frontiers and food security among Maasai pastoralistsin Kenya. In Bruins, H. J., and Lithwick, H. (eds.), The Arid Frontier, Kluwer Academic,Netherlands, pp. 185–208.

Sindiga, I. (1992). The future of Maasai pastoralist. Professional Geographer 44(1): 101–102.Spear, T. (1993). Introduction. In Spear, T., and Waller, R. (eds.), Being Maasai: Ethnicity and

Identity in East Africa, James Curry, London.Spear, T. (1993). Being “Maasai,” but not “People of Cattle.” In Spear, T., and Waller, R. (eds.),

Being Maasai: Ethnicity and Identity in East Africa, James Curry, London.Spear, T., and Waller, R. (eds.). (1993). Being Maasai: Ethnicity and Identity in East Africa,

James Curry, London.Spencer, P. (1988). The Maasai of Matapato: A Study of Rituals of Rebellion, Manchester Uni-

versity Press, Manchester.

P1: GVM/GDX



Swift, J. (1986). The economics of production and exchange. In Adamu, M., and Kirk-Greene,A. (eds.), Pastoralists of the the West African Savanna, Manchester University Press,Manchester.

Swift, J. (ed.). (1990). Providing Services for Nomadic People: A Review of the Literature andAnnotated Bibliography, UNICEF, New York.

Talle, A. (1987). Women as heads of houses: The organisation of production and the role ofwomen among pastoral Maasai in Kenya. Ethnos 1–2: 50–80.

Talle, A. (1994). The making of female fertility: Anthropological perspectives on a bodily issue.Acta Obstetrica Gynecologica Scandinavica 73: 280–283.

Talle, A. (1999). Pastoralists at the border: Maasai poverty and the development discourse inTanzania. In Anderson, D. M., and Broch-Due, V. (eds.), The Poor Are Not Us: Povertyand Pastoralism in Eastern Africa, James Curry, Oxford.

Thompson, D. M. (ed.). (1997). Multiple Land-Use: The Experience of the Ngorongoro Conser-vation Area Tanzania, IUCN, Gland, Switzerland.

UNDP (1999). Human Development Report, Oxford University Press, Washington.von Mitzlaff, U. (1988). Maasai women: Life in a patriarchal society: Field research among the

Parakuyo, Tanzania Trickster Verlag: Munich pp. 166.World Bank. (2001). Retrieved on 5 February, 2001, from: http://www.worldbank.org/

participation/kenyacase.htm.

maasai socioeconomic conditions: a cross-border comparison

Documents