cattle commercialization in rural south africa: livelihood...
Post on 19-Apr-2018
220 Views
Preview:
TRANSCRIPT
© Kamla-Raj 2014 J Hum Ecol, 45(3): 207-221 (2014)
Cattle Commercialization in Rural South Africa:Livelihood Drivers and Implications for Livestock
Marketing ExtensionJorine T. Ndoro1*, Maxwell Mudhara2 and Michael Chimonyo3
1Department of Agricultural Extension and Rural Resources Management, 2African Centre forFood Security, 3Department of Poultry and Animal Sciences, University of KwaZulu-Natal,
P/Bag X01 Scottsville, 3209, Pietermaritzburg, South Africa*E-mail: jorinendoro@gmail.com
KEYWORDS Market Participation. Sustainable Livelihood Framework. Double-Hurdle Model. AgriculturalExtension. South Africa
ABSTRACT Commercialization of livestock farming systems remains a challenge in rural South Africa. Recentempirical evidence places agricultural extension at the forefront of policy strategy to address this challenge. Thisstudy applies the sustainable livelihood framework (SLF) to quantitatively analyze the factors confoundingparticipation in cattle markets for the purpose of informing agriculture extension programming. Based on a datasetcompiled from a household survey of 230 randomly selected smallholder cattle farmers in Okhahlamba LocalMunicipality (OLM), a Double-Hurdle econometric estimation technique is used to determine factors within theSLF influencing market participation and supply volumes decisions. The results reveal that the low rate of marketparticipation could be explained by the broader aspects of livelihoods of smallholder cattle farmers, includinglimited access to financial, social and natural capital, as well as the difference in livelihood strategies and motivations.Based on these findings, the study draws the implications for the design of livestock extension programs in OLM,and South Africa in general.
INTRODUCTION
Background of the Study
Market participation is an important ingredi-ent for agricultural and rural development. Com-mercialization of smallholder farming systemsthrough active participation in cattle markets hasthe potential to exploit developing regions’ com-parative advantages and transform rural econo-mies (Boughton et al. 2007; Rios et al. 2009;Mathenge et al. 2010). Commercializing small-holder farming systems leads to increased pro-ductivity and improved quality of produce, there-by contributing to improved incomes. Hence,market participation by smallholder cattle farm-ers has the potential to lead to specialized, mar-ket-oriented farming systems (Rios et al. 2009).
In South Africa, the recent growth in live-stock markets brought about by high popula-tion and income growths, urban migration, glo-balization, and their associated changes in life-styles and consumer preferences, has present-ed new opportunities for smallholder livestockfarmers to be integrated into the market econo-my (Delgado et al. 2001; Coetzee et al. 2005;Uzchezuba et al. 2009). Cattle production con-tributes between 25% and 30% per annum to
South Africa’s national agricultural GDP(Musemwa et al. 2008). In addition to its impor-tance in the national economy, cattle produc-tion is a key livelihood strategy of the resource-poor smallholder farmers in South Africa, wherearound 40% of the total cattle herd size is ownedby communal and emerging farmers (NationalDepartment of Agriculture 2011). Cattle produc-tion by smallholder farmers constitutes a majorlivelihood strategy particularly for farminghouseholds living in marginal areas with degrad-ed lands, and meager economic opportunities,and hence acute poverty, food insecurity andunemployment (Machethe 2004).
The appeal of cattle as a viable agriculturalinvestment option has influenced rural devel-opment policies in South Africa. Several strate-gic intents have been devised to transform therural livestock sector towards a commercialisedindustry. The National Livestock DevelopmentStrategy proposes to support smallholder andemerging farmers to be competitive and profit-able (National Department of Agriculture 2006).The strategy proposes to support smallholderlivestock farmers through creation of an enablingpolicy environment, investment in rural commer-cial and cooperative infrastructure, market de-velopment, training and research, and equitable
208 JORINE T. NDORO, MAXWELL MUDHARA AND MICHAEL CHIMONYO
participation, and integration into sustainablerural development (National Department of Ag-riculture 2006). In addition, for the livestock sec-tor, the agricultural marketing strategy has setout to develop commodity groups/associationsfor ease of smallholder farmers’ access to marketinformation and agricultural marketing infrastruc-ture (National Department of Agriculture 2010).These incentives have opened up a variety ofmarket channels for livestock farmers, includingauctions, speculations, abattoirs, butcheries, aswell as farm-gate sales (Nkosi and Kirsten 1993).
Despite the congruence of incentive struc-tures and processes, the cattle markets in SouthAfrica remain characterized by low participationrates of smallholder farmers. Recent studiesfound the levels of cattle commercialization tobe directly proportional to the holding, with ratesof 33% for herd of 10 or less cattle, 52% for 11-20cattle owners, and 85% for 20 or more cattle keep-ers (Coetzee et al. 2005; Lehloenya et al. 2007;Musemwa et al. 2007; Groenewald and Jooste2012). These studies have documented off-takerates ranging between 5% and 10 % among com-munal lands/smallholder farmers compared to 25% for commercial farmers (Musemwa et al. 2010).
As studies in the agricultural economics lit-erature explain, lower levels of smallholder farm-ers’ participation in agricultural markets can beexplained by the incidence of costs or/and non-commercial motives. The transaction costs con-sist of fixed transaction costs arising from im-perfect information, such as the search cost forcustomers with good terms and conditions, ne-gotiations and bargaining, screening, enforce-ment, and the costs proportional to the level ofactivity encompassing per unit costs of marketaccess such as transportation and imperfect in-formation (Key et al. 2000). As Barrett (2008)explains, the extent of these costs largely de-pends on the household’s capability, as definedby its endowment including education, physi-cal infrastructure, social networks and access topublic goods such as agricultural extension ser-vices, roads, information and communication.Also, there is another body of literature con-tending that in southern African, cattle are keptfor wealth storage rather than income genera-tion (Doran et al. 1979). The asset accumulationand ownership benefits such as security andprestige outweigh market incentives (Jarvis1980).
Agricultural extension is one aspect thatshould be strengthened to reduce the transac-tion costs faced by smallholder farmers in thelivestock markets (Bahta and Bauer 2007; Uche-zubal et al. 2009). This realization is emergingeven as agricultural extension approaches areundergoing paradigm shifts, from a top-down,technology transfer model of extension deliveryto multifunctional, farmer-centered, participatoryand systems–based approaches to rural devel-opment (Duvel 2000; Coetzee et al. 2005; Anan-dajayasekeram et al. 2008; Swanson and Rajalah-ti 2010). The new approaches are supposed toaddress the real needs of the farmers and en-courage their innovativeness.
Research Problem and Objective
Empirical studies in agricultural marketing inSouth Africa consider agricultural extension onlyas a discrete ingredient whose access can offsetor moderate high transaction costs and otherchallenges. Such studies do not explore the vary-ing degrees to which agricultural extension canplay an integrative role that fosters agriculturalmarket participation. This leaves a vacuum inthe understanding of the relevance of differentagricultural extension models and methods inaddressing the complexity of farmers’ issues re-lated to livestock markets participation.
For a more practical approach to analyzingcattle commercialization and addressing the roleof extension in livestock market participation,an application of the sustainable livelihood frame-work (SLF) has two unique advantages. First,the framework gives an explicit consideration ofboth aspects of challenges and barriers to mar-ket participation, that is, transaction cost andfarmers’ motivations (Department for Interna-tional Development 1999). Second, it offers anintegrative programming framework for povertyalleviation in a sustainable manner (Krantz 2001).In line with appropriate extension models forSouth Africa (Duvel 2000), the SLF is a respon-sive and participatory programming frameworkthat builds on people’s strengths, and at thesame time attempts to overcome their challeng-es and barriers at multiple levels, thus ensuringthat micro-level challenges inform policy devel-opment and macro-level environment enablespeople to build on their strengths (DFID 1999).
Leveraging on this appeal, the objective ofthe study is to empirically investigate the ef-
CATTLE COMMERCIALIZATION IN RURAL SOUTH AFRICA 209
fects of factors under different SLF componentson market participation decisions among small-holder cattle farmers, for the purpose of recom-mending appropriate agricultural extension mod-els and methods.
The remainder of this paper is sub-dividedinto five sections. The subsequent section over-views the key findings of previous empirical stud-ies in the domain of livestock market participa-tion in South Africa. It is followed by a method-ological section outlining the empirical strategyadopted by the study, and a section discussingthe empirical findings. The last two sectionsconclude and draw the implications for agricul-tural extension in South Africa.
Literature Review
Market participation cannot be explained bya single factor (such as price incentives) since itstands out to be both a consequence and a causeof development (Barrett 2008). Farm households’market participation requires access to technol-ogy, private and public (institutional and phys-ical) productive assets, which entails varioussunk and fixed costs, coordination problems, andliquidity constraints at all decision-making lev-els (Barrett 2008). The costs associated withmarket transactions (customer search, negotia-tion, bargaining, screening, etc.), in particular,determines difference in market relations amongsmallholder farmers (Key et al. 2000). These trans-action costs, which are largely dependent onhousehold-specific factors, can push smallhold-er farmers’ livelihoods into low-level equilibriumtraps of semi-subsistence farming systems (Dor-ward et al. 2003; Barrett 2008). In this line ofanalysis, a number of empirical studies in thedomain of cattle market participation withinSouth Africa have documented various con-founding livelihood factors. This section pre-sents an overview of these findings, with a sus-tainable livelihoods lens.
On the livelihood vulnerability context, cat-tle mortality and thefts were found to be signif-icant factors explaining positive livestock mar-ket participation decisions in Limpopo, EasternCape and Northwest Provinces (Montshwe2006).
With regard to livelihoods assets, empiricalstudies have documented the significance ofhuman, physical, financial, and natural capital.With regard to human capital, Makhura (2001)
found that female-headed-households are morelikely to participate in the Northern Province’slivestock market. In the Northern Cape Province,Uchezubal et al. (2009) found that householdswith few and experienced members have highchances of engaging in livestock markets, where-as shorter distances to market and market infra-structure enhanced participation. The findingthat smaller household sizes could explain pos-itive market participation decisions was sharplycontrasted by Monthswe (2006) who revealedthat, in Limpopo, Eastern Cape and NorthwestProvinces, larger households in terms of thenumber of members are more likely to participatein livestock market. He further found that trainedfarmers and those who live within shorter dis-tances to market had more probabilities of par-ticipating in livestock markets. These resultswere vindicated by Bahta and Bauer (2007) inthe Free State Province. The significance of ac-cess to extension services has been commonlyevidenced in the literature, including studiessuch as Uchezubal et al. (2009) and Bahta andBauer (2007). Access to market information isalso constantly revealed as a key market partic-ipation factor (Bahta and Bauer 2007).
The endowment in natural resources, partic-ularly the herd sizes, has also been found toinfluence market participation rates among small-holder livestock farmers (Makhura 2001; Mont-shwe 2006; Bahta and Bauer 2007). The signifi-cance of the influence of financial assets hasalso been documented. For instance, indebted-ness was found to be a significantly negativefactor of market participation among small-scalelivestock farmers in the Northern Cape (Uche-zubal et al. 2009).
On the structures and processes, however,important processes in the livestock industryhave been overlooked by empirical studies. Yet,qualitative studies have pointed out that com-pliance with livestock management regulations(such as the Livestock Identification Act) fig-ures among livestock marketing constraints facedby small-scale farmers in South Africa (Coetzeeet al. 2005; Groenewald and Jooste 2012).
The motivational aspect of livestock market-ing has also been investigated. Non-commercialmotives in the keeping of livestock include eco-nomic functions (for example, wealth storage),agro-economic functions (for example, provisionof draft power), agro-ecological functions (forexample, provision of manure), nutritional (for
210 JORINE T. NDORO, MAXWELL MUDHARA AND MICHAEL CHIMONYO
example, provision of milk) as well as socio-cul-tural functions (for example,, dowry) (Nkosi andKirsten 1999; Lehloenya et al. 2007; Groenewaldand Jooste 2012). Makhura (2001) showed thatthe more unearned incomes (pension) the house-hold receives, the less the probability of its mar-ket participation, suggesting the predominanceof non-commercial motives. However, these find-ings were in contrast with the findings of Mont-shwe (2006) showing that, in the Limpopo, East-ern Cape and Northwest Provinces, householdwho received unearned incomes had morechances of participating in livestock markets.
Notwithstanding the important insight ofthese studies, none of them analyse transactioncost and motivational aspect in an integratedmanner, which the objective of this study.
METHODOLOGY
Study Area
This study was conducted in OkhahlambaLocal Municipality (OLM), a 344,000ha munici-pality in the UThukela District of the KwaZulu-Natal Province (see Fig. 1). The 2007 populationcensus indicates that the municipality is inhab-ited by 151,414 people (or 28,508 households),mainly traditional households (56%), illiterate(38%), and communal lands dwellers (OLM 2012).Vast majorities of these people are deprived ofpublic infrastructure (with only 39%, 63%, and44% having access to electricity, water in theirdwellings, and transportation, respectively)(OLM 2012). As reported by the municipality,the harsh economic conditions are such thataround 36% of household do not receive anyincome, whilst 37% earn less than R9,600 (aroundUS$1,100) per annum (OLM 2011).
In this area, commercial and subsistence farm-ing coexist, although geographically separated(a legacy of the segregationist regime of theapartheid era). Smallholder farmers, mainly en-gaging in maize, vegetable, and livestock pro-duction, occupy the marginal areas, mainly thefoothills of the Drakensberg mountain rangechain, characterized by low-fertility lands (Elle-boudt 2012). Although only 22% of the econom-ically active population engages in crop farming(OLM 2012), 55% of households living on com-munal land engage in livestock farming, mainlyconsisting of cattle, goats and sheep (Elleboudt2012). Mixed livestock-crop farming system is a
special feature of agriculture in the foothills ofDrakensberg region, where grazing is scheduledsuch that cattle are sent to uphill areas duringthe cropping season in summer, while all the landbecomes grazing land off-season in winter (Elle-boudt 2012). This creates overstocking tenden-cies among locals with the associated environ-mental consequences, and the situation is rein-forced by the lack of property rights and en-forcement mechanisms such as fencing. The areais also know to experience harsh climatic condi-tions, characterized by an interchange ofdroughts conditions in summer and heavy snowin winter, making the palatability of the naturalgrasslands very seasonal, and farmers have toprovide supplementary feeding (Elleboudt 2012).
Empirical Framework
Following the prescriptions of Bellemare andBarrett (2006) and other previous studies suchas Winter-Nelson and Temu (2005) and Alene etal. (2008), this study uses a sample selectionmodel to unpack market participation behavioramong smallholder farmers. Hence, to estimatethe influence of livelihood factors in explainingparticipation and supply decisions among cat-tle farmers, this study adopts the Double-Hur-dle (DH) econometric technique, as initially pro-posed by Cragg (1971).
Under this empirical strategy, a cattle farmerhas to cross two hurdles to become a partici-pant in a cattle market. First, the farmer becomesa “potential participant” after crossing the firsthurdle, i.e. after making a positive decision toparticipate in the livestock market. A potentialparticipant, capability factors will determine hisactual/observed level of participation in the sec-ond hurdle. Therefore, the DH model is a two-equation framework (Matshe and Young 2004;Moffatt 2005; Ground and Koch 2008), as de-picted in the equation 1.
Considering *iI as a binary choice variable,
*siQ as a latent variable which reflects the num-
ber of cattle sold (therefore the observed vari-able, Qi, being determined as ** s
iii QIQ ), Z and ábeing the vectors of factor explaining the deci-sion of participation and their influences respec-tively, and X and â being the vector of factorsexplaining the intensity of participation and theirinfluences respectively; the DH model can bewritten as follow (Matshe and Young 2004;Moffatt 2005; Ground and Koch 2008):
CATTLE COMMERCIALIZATION IN RURAL SOUTH AFRICA 211
Fig.
1.
Lan
d us
e m
ap o
f th
e O
khah
lam
ba L
ocal
Mun
icip
alit
y sh
owin
g di
p ta
nks
Sour
ce: A
utho
rs -
bas
ed o
n la
nd c
over
dat
aset
pro
vide
d by
the
Eze
mve
lo K
waZ
ulu-
Nat
al W
ildlif
e (h
ttp://
ww
w.b
gis.s
anbi
.org
/kzn
/land
cove
r.asp
).
212 JORINE T. NDORO, MAXWELL MUDHARA AND MICHAEL CHIMONYO
where,The log-likelihood function for the DH mod-
el is:
The analysis of marginal effect helps to as-sess the impact of the exogenous variables onthe dependent variables. To do so, the uncondi-tional mean is decomposed into the effect onthe probability of participating and the effect onthe conditional level of participation and differ-entiating these components with respect to eachexplanatory variable. The unconditional meancan be written as:
The probability of participation and the ex-pected number of cattle sold conditional on par-ticipation are:
and
Sampling and Data Collection
The above-outlined model was fed withhousehold-level data collected in two phases.The first phase, the researchers conducted par-ticipatory rural appraisals (PRA) during the pe-riod of June to October, 2012. Over the course ofthis period, key informant interviews with exten-sion personnel were conducted, followed byfocus group discussions with knowledgeablemembers of various dip-tank users associations(DUAs), through their mother cooperative, theOkhahlamba Livestock Cooperative (OLC). Thisphase was meant to picture key challenges andbarriers around cattle marketing in OLM, as per-ceived by OLC members1. The information gath-ered during this phase was used to device astructured household survey questionnaire thatwas pilot-tested and administered by trainedfield enumerators during the second phase, span-ning from November, 2012 to February, 2013. Farmhouseholds were randomly selected for struc-tured interviews based on a two-stage randomsampling technique. In the first stage, 12 out of
31 DUAs were randomly selected using simplerandom selection technique, that is, with a ran-dom number generator. In the second stage, 20members of each pre-selected DUA were select-ed randomly based on a systematic random sam-pling procedure. In total, 230 heads of cattle farmhouseholds were interviewed at their home-steads.
The information gathered during the surveywas on the various livelihood characteristics ofthe farm household, based on various compo-nents of the SLF, including the vulnerability con-text, the livelihood assets, structures and pro-cesses, livelihoods strategies, and livelihoodoutcomes. Furthermore, the questionnaire cap-tured key market participation behaviors, as wellas farmers’ perceptions about market participa-tion challenges and barriers.
Empirical Estimation and Sample Description
To estimate the effect of livelihood factorson the market participation decisions (first hur-dle), the Probit regression was used. The inten-sity of participation levels, the second stage (sec-ond hurdle) was estimated using a truncated re-gression model (Wooldridge 2002). Prospectivevariables were first shortlisted based on the in-formation gathered during the PRA phase as wellas key factors unveiled by previous empiricalstudies. Thereafter, a prospective variable wasselected for the regression based on the signif-icance of its contribution to the improvement ofthe model’s fit that is, the Log-Likelihood ratio(LR) test (Wooldridge 2002). This techniqueguaranteed that the selected variables would givethe best fit.
Multicollinearity was tested using a correla-tion matrix presented in Table 4 (appendix), theresults of which suggest that multicollinearitywas not a serious problem in the data. To curbthe potential heteroskedasticity in the model,the study used the heteroskedasticity-robuststandard errors for parameter estimates.
For the model on the first hurdle, that is, thedecision to participate in livestock markets, theself-selection bias was corrected for each par-ticipating household by generating the InverseMills Ratio (IMR) from the predicted probabili-ties of the probit model and subsequently in-cluding it as an explanatory variable in the trun-cated regression of level of participation (Wool-dridge 2002). Although theory does not point to
hurdle second hurdlefirst
'*
'*
iisi
iii
XQZI
20
01,
00
~
N
i
i
)(1)(ln)(1ln
''
0
'' ii
ii
iXYZXZLogL
(1)
(2)
)0|()0(]|[ iiii YQEQPXQE
(3)
'
' )()0( iii
XZQP
ii
ii
ii
i
i
iii dY
XQT
Y
QXQQE
0
'
22
1' )(
1)0|(
(4)
(5)
CATTLE COMMERCIALIZATION IN RURAL SOUTH AFRICA 213
the need of imposing exclusion restrictions inthe Double-Hurdle model as with the Heckmanmodel, an exclusion restriction was imposed inthe model since the IMR variable can be corre-lated with the vector of explanatory variables inthe intensity model especially if both hurdleshave equal vectors of explanatory variables(Wooldridge 2002). It is recommended that avariable that is likely to affect the selection butnot have partial effect on the intensity modelcan conveniently be excluded. Potential factorsto be excluded in the intensity model were thosethat explained, to some extent, the fixed transac-tion costs, since they influenced only the firstparticipation decision model (Key et al. 2000;Alene et al. 2008). Using the LR test, distance tomarket was excluded.
In consideration of the above-mentionedtechnicalities, and following the design of theSLF, the sample statistics describing the vari-ables used in the empirical model are presentedin Table 1. As the table shows, nearly 48% ofinterviewed farm households had participatedin cattle markets. Vulnerability to climatic condi-tions was more pronounced among the inter-viewed households, with staggering proportionsof 40-50% and 20-26% reporting cattle loss fol-lowing severe winters and droughts, respective-ly. However, the rates were not statistically dif-ferent between market participants and non-par-ticipants groups. With regard to human capitalindicators, the majority of interviewed house-hold heads were born in between 1953 and 1955,and household heads in the market participantsgroup were significantly younger than their coun-terparts. The majority of interviewed farmers (50-65%) had recently benefited from extension train-ing organized by the provincial Department ofAgriculture. Households in market participantsgroup, however, had relatively and significantlyreceived more T&Vs than farmers in the non-participants group.
Regarding social capital, the majority of in-terviewed farmers were members to the Okhahl-amba Livestock Cooperative (OLC), although therate of membership was significantly higher inthe market participants group. The rates of par-ticipation in farmer-to-farmer extension programswere, however, lower, ranging from 12% in theparticipants group to 6% in the non-participantsgroup. The difference in the participation ratesbetween groups was however not statisticallysignificant. The rates of participation in the sav-
ings groups (stokvels) ranging from 34% (non-participants groups to 42% (participants) wererecorded.
For physical capital, about 11% of inter-viewed households owned a tractor and stayedwithin 21 km distance to the Dukuza cattle mar-ket. With regard to natural capital, the averageherd sizes ranged between eight for non-partic-ipants, and 14 for participants, and the differ-ence was statistically significant. The walkingdistance to the nearest source of water was sta-tistically different between the two groups, withthe majority of households in the participantsgroup walking 20 min, while their counterpartswalk 14 minutes. With regard to the transform-ing structures and processes, the majority inboth groups had tagged their cattle althoughthe participating group recorded a significantlyhigher tagging rate.
For the livelihood outcomes, incomes fromlivestock sale were significantly more importantin the livelihoods portfolio among interviewedmarket participant than in the non-participantgroup, whereas the importance of remittanceswas significantly more pronounced amonghouseholds in the non-participants group. Theexpected cattle price was about R5, 500.
RESULTS AND DISCUSSION
The results of the participation and supplymodels are presented in Table 2 and Table 3. Thevariables are presented based on the compo-nent of the SLF to facilitate their discussion.
Livelihood Assets
The results of both regressions show thatamong human and social capital factors, the co-efficient of farmer-to-farmer extension variableis significant for the supply model. This findingallows inference that, given positive participa-tion decision, potential participants that receivedextension trainings and information sharing ses-sions through their groups tend to supply morecattle to the market. Therefore, this result sug-gests that farmers do capitalize on the informa-tion networks when deciding the amount of cat-tle to be sold. This finding vindicates the con-tention that what matters for positive economicoutcomes among the poor is not membership ingroups, but the quality and quantity of resourc-es (information) flowing within those networks(DFID 1999; Kirsten et al. 2009).
214 JORINE T. NDORO, MAXWELL MUDHARA AND MICHAEL CHIMONYO
On household’s financial capital, participa-tion in saving groups turns out to be a majorpredictor of the decision to participate as a cat-tle seller. Other livelihood factors remaining un-changed, opening an account in a local savinggroup (or a stokvel) significantly increases theprobability of participating in cattle market as aseller by 13.4%. Arguably, smallholder farmersthat belong to saving groups have access tocredit that enables them to increase the produc-tivity and market value of their herd, therebyincreasing the prospect of market participation.
On natural capital, the regression results in-dicate that cattle market participation and sup-ply decisions are significantly and positivelygoverned by the cattle herd size. Adding onecattle to the herd significantly increases thechances of participating in cattle market as aseller by 1.8%, ceteris paribus. These findingsvindicate the hypothesis that agricultural mar-ket participation is associated with its produc-tivity (Lapar et al. 2003; Rios et al. 2009) and theempirical evidence that shifting to commercialcattle farming systems in Southern Africa re-quires growth in herd sizes (Behnke 1987).
The results of the Probit model also showthat the cattle breed has a significant effect onmarket participation decisions. All other factorsin the model remaining constant, shifting froman exclusively indigenous breed (Nguni) herdto mixed/crossbred herds, towards exotic breedsignificantly reduces the farmer’s prospect ofcattle market participation, implying that farm-ers who keep indigenous breed are more likelyto participate in market as sellers. This suggeststhat farmers do take into account the breed whendeciding to sell their cattle. This is probably dueto the fact that the indigenous breed of the east-ern and northern South Africa is more fertile,matures earlier, is well adapted to low qualityfeed, and therefore easily replaceable comparedto other breeds (Bayer et al. 2004). This findingsuggests that both the quantity and quality ofcattle herd are important for a pro-poor marketdevelopment strategy.
Lastly, the coefficient of walking distance tothe nearest water source also has a significantpositive effect, and it can be inferred that OLMcattle farmers staying far from water sources suchas rivers and dams have more chances of partic-ipating in cattle market as sellers. This findingsignals the potential of distress sales amongsmallholder farmers, particularly during adverseperiods of prolonged drought (Elledoubt 2012).
Transforming Structures and Processes
The empirical results of the participationmodel yield a positive and significant coefficientfor cattle tagging. These results suggest thatcompliance with the Livestock Identification Actis a key factor in cattle marketing, and perhapsthe most important constraint to the cattle mar-ket participation in the empirical model. Ceterisparibus, registering (branding and marking) thecattle herd increase market participation propen-sity by 21%. This finding vindicates the asser-tions of Coetzee et al. (2005) and Groenewaldand Jooste (2012) that livestock registration leg-islation is an important factor for a pro-poor cat-tle market development policy in South Africa.
Livelihood Strategies
The results of the participation model showthat the coefficient of the rank of cattle income,among other income generation strategies, isnegative and significant. This result means thathouseholds whose primary income earner is cat-tle farming are more likely to participate in cattlemarket, suggesting that the portfolio of liveli-hood strategies explains the differences in cat-tle market participation rates among smallholderfarmers. This implies that the degree of special-ization in cattle farming is an important predictorof cattle market development.
Livelihood Outcomes
The results in Table 2 show that the coeffi-cient of the rank of remittances in the incomeportfolio is significantly positive, suggestingthat cattle farmers who regularly secure moreunearned incomes such as remittances from theirfamily members and friends are not likely to par-ticipate in cattle market. This result is in line withthe walking-bank hypothesis of livestock mar-keting (Bellemare and Barrett 2006), suggestingthat market participation decisions are drivenby the need to cater for immediate householdneeds when cash is not otherwise available.
The results also show that the coefficient ofthe expected price variable is only significant inthe supply model. Consistent with the findingsof previous studies done in developing coun-tries such as the study by Alene et al. (2008),this empirical finding reveals that smallholderfarmers do not necessarily consider informationon prevailing price incentive when deciding to
CATTLE COMMERCIALIZATION IN RURAL SOUTH AFRICA 215Ta
ble
1: D
escr
ipti
on o
f va
riab
les
and
t-te
st o
f eq
ualit
y of
mea
ns f
or i
ndep
ende
nt v
aria
bles
use
d in
the
em
piri
cal
mod
el
Vari
able
cat
egor
y,Va
riab
le d
escr
iptio
nM
easu
rem
ent
Valu
e la
bels
Part
icip
ants
Non
- pa
rtic
ipan
tsp-
valu
eSL
F co
mpo
nent
,
N=
113
N=
117
and
vari
able
nam
e
Dep
ende
nt v
aria
bles
MA
RK
PAR
TD
ecis
ion
to p
artic
ipat
e in
cat
tleD
umm
y1
= th
e ho
useh
old
has
mar
ket
duri
ng t
he l
ast
thre
e ye
ars
part
icip
ated
in
cattl
em
arke
t as
a s
elle
r, an
d0
= ot
herw
ise.
TO
TSO
LD
Num
ber
of c
attle
sol
d ov
er t
hepe
riod
of
the
last
thr
ee y
ears
Cou
ntIn
depe
nden
t va
riab
les
1. V
ulne
rabi
lity
cont
ext
SNO
WLO
SSE
xper
ienc
e w
ith c
attle
dea
ths
Dum
my
1= t
he h
ouse
hold
.49
.41
.246
resu
lting
fro
m h
eavy
sno
w o
ver
the
expe
rien
ce c
attle
dea
thla
st t
hree
yea
rsat
trib
utab
le t
o he
avy
snow
ove
r th
e la
st t
hree
year
s; 0
= ot
herw
ise.
DR
OU
GH
TLO
SSE
xper
ienc
e w
ith c
attle
dea
ths
Dum
my
1= t
he h
ouse
hold
exp
erie
nce
.20
.26
.273
resu
lting
fro
m p
rolo
nged
per
iod
ofca
ttle
deat
h at
trib
utab
ledr
ough
ts o
ver
the
last
thr
ee y
ears
to d
roug
ht c
ondi
tions
ove
rth
e la
st t
hree
yea
rs;
0= o
ther
wis
e.As
set
pent
agon
Hum
an c
apita
lB
IRT
HD
AYH
HH
Yea
r of
bir
th o
f th
e he
ad o
f ho
useh
old
Con
tinu
ous
1955
.94
1953
.08
.076
GO
VEX
TV
ISIT
Acc
ess
to
exte
nsio
n tr
aini
ng a
ndD
umm
y1=
the
far
mer
rec
eive
d a
.646
0.5
043
.030
visi
t fr
om g
over
nmen
t ex
tens
ion
gove
rnm
ent
exte
nsio
nag
ents
ove
r th
e la
st t
hree
yea
rstr
aini
ng o
r vi
sit
over
the
last
thr
ee y
ears
;0=
othe
rwis
e.So
cial
cap
ital
OL
CM
EM
BM
embe
rshi
p in
OLC
Dum
my
1= t
he h
ead
of h
ouse
hold
.84
.74
.071
is a
mem
ber
of O
LC;
0= o
ther
wis
e.F2
FEX
TA
cces
s to
fa
rmer
-to-
farm
erD
umm
y1=
the
far
mer
rec
eive
d.1
239
.068
4.1
54ex
tens
ion
trai
ning
s ov
er t
hefa
rmer
-to-
farm
erla
st t
hree
yea
rsex
tens
ion
trai
ning
or
info
rmat
ion
sess
ions
ove
rth
e la
st t
hree
yea
rs;
0=ot
herw
ise.
Fina
ncia
l ca
pita
lSA
VG
RO
UP
If t
he h
ouse
hold
is
a m
embe
rD
umm
y1=
The
hea
d of
hou
seho
lds
.42
.34
.198
of a
sto
kvel
ave
mon
ey i
n a
stok
vel;
0= o
ther
wis
e.
216 JORINE T. NDORO, MAXWELL MUDHARA AND MICHAEL CHIMONYOTa
ble
1: C
ontd
...Va
riab
le c
ateg
ory,
Vari
able
des
crip
tion
Mea
sure
men
tVa
lue
labe
lsPa
rtic
ipan
tsN
on-
part
icip
ants
p-va
lue
SLF
com
pone
nt,
N
=11
3
N
=11
7an
d va
riab
le n
ame
Phys
ical
cap
ital
TR
AC
TO
RO
wne
rshi
p of
tra
ctor
–an
indi
cato
rD
umm
y1=
The
hea
d of
hou
seho
ld.1
0.1
1.7
34of
fix
ed c
attle
pro
duct
ion
cost
owns
a t
ract
or;
0=ot
herw
ise.
RD
DIS
TT
OD
UK
UZ
ASh
orte
st d
rivi
ng d
ista
nce
(Km
)C
onti
nuou
s21
.577
020
.265
0.4
61fr
om t
he c
omm
unity
’s d
ip t
ank
to D
ukuz
a ca
ttle
mar
ket
plac
e (
mea
sure
d us
ing
GPS
nav
igat
ion
softw
are)
– a
n in
dica
tor
of f
ixed
cattl
e tr
ansa
ctio
n co
stN
atur
al c
apita
lH
ER
DSI
ZETo
tal
num
ber
of c
attle
ow
ned
atC
ount
14.6
88.
92.0
00th
e tim
e of
int
ervi
ew –
– a
nin
dica
tor
of m
anag
ed e
cosy
stem
good
s an
d se
rvic
esC
ATT
LE
BR
EE
DTy
pe o
f br
eed
owne
d by
the
far
mer
Cat
egor
ical
–1=
Ngu
ni;
2= M
ixed
;1.
731.
77.5
78 –
an
indi
cato
r of
man
aged
Ord
inal
3= E
xotic
bre
edec
osys
tem
goo
ds a
nd s
ervi
ces
WAT
ERSO
UR
CED
IST
Wal
king
dis
tanc
e (i
n m
inut
es)
Con
tinu
ous
21.6
114
.74
.097
betw
een
the
hous
ehol
d an
d th
ene
ares
t ca
ttle
wat
er s
ourc
e –
an i
ndic
ator
of
acce
ss t
o na
tura
lec
osys
tem
ser
vice
s.Tr
ansf
orm
ing
Proc
esse
san
d st
ruct
ures
CAT
TL
ETA
GC
attle
are
tag
ged
- an
ind
icat
orD
umm
y1=
Cat
tle c
onfo
rms
to t
he.9
5.8
5.0
20of
com
plia
nce
wit
h L
ives
tock
requ
ired
ide
ntif
icat
ion
Iden
tifi
cati
on A
ctta
gs,
0= o
ther
wis
e.Li
velih
ood
stra
tegi
esC
ATT
LE
INC
RA
NK
The
im
port
ance
(ra
nk)
of i
ncom
esC
ateg
oric
al1
= m
ost
impo
rtan
t3.
892.
86.0
00fr
om c
attle
sal
es a
mon
g re
gula
r ~
5=
leas
t im
port
ant
sour
ces
of e
arne
d in
com
e L
ivel
ihoo
d ou
tcom
esR
EM
ITR
AN
KT
he
impo
rtan
ce (
rank
) of
Cat
egor
ical
1 =
mos
t im
port
ant
2.65
3.44
.000
rem
ittan
ces
amon
g re
gula
r so
urce
s~
5= l
east
im
port
ant
of u
near
ned
inco
me
EX
PPR
ICE
Exp
ecte
d ca
ttle
pric
e at
far
m g
ate
Con
tinu
ous
5480
.767
055
95.0
095
.339
and
from
spe
cula
tors
in
the
com
mun
ity
Inve
rse
mill
s ra
tio (
IMR)
The
sta
ndar
d no
rmal
pro
babi
lity
Con
tinu
ous
dist
ribu
tion
func
tion
over
the
stan
dard
nor
mal
cum
ulat
ive
dist
ribu
tion
func
tion
of t
hepr
edic
ted
prob
abili
ties
Sour
ce:
Aut
hors
’ su
rvey
, 20
12-2
013
CATTLE COMMERCIALIZATION IN RURAL SOUTH AFRICA 217
sell their cattle. Nonetheless, the evidence sug-gests that given positive participation decisions,smallholders will consider price signals whendeciding upon the number of cattle to be soldon the market. These results possibly suggestthat market participation and volume decisionsare not taken simultaneously, that is, althoughpredisposed to selling their cattle, livestock farm-ers do not pre-commit the number of cattle to besold before learning information about the pre-vailing market conditions, especially the price.However, the negative sign indicates that thereis a considerable scope of non-commercial moti-vations such as wealth storage, and cattle isonly sold when a farm household faces press-ing cash needs (Doran et al. 1979; Groenewaldand Jooste 2012). The results validate the find-ing that farmers who manage to secure more in-comes from alternative sources are less likely toparticipate in cattle market.
CONCLUSION
This study falls within the realm of participa-tory agricultural extension approaches. Its pur-
pose is to investigate the livelihood drivers ofsmallholder cattle farmers’ participation and sup-ply decisions in OLM within a framework of in-forming the design of livestock extension archi-tecture in South Africa. The goal is to explainthe low rates of cattle market participation bysmallholder farmers in the region and suggestappropriate extension models for extension pro-gramming in the context of OLM. Based on ahousehold survey dataset compiled from a sur-vey of 230 randomly selected cattle farm house-holds, the study runs a Double-Hurdle econo-metric model to calibrate the effect of factorsunder various components of the SLF on theparticipation and supply decisions.
The empirical results of this study reveal thatthe low rate of market participation cannot besimply explained by endowments and accessfactors (the determinants of transaction cost),but the broader aspects of livelihoods of small-holder cattle farmers in South Africa. Notably,this study finds that the difference in access tofinance, natural capital endowments, and liveli-hood strategies could explain the rate of market
Table 2: Livelihood determinants of farmers’ decisions to participate in cattle market in OLM, Resultsestimated using Probit regression model
SLF component and variable name Coefficient Marginal effects p>|Z|Dependent variable: MARKPART1. Vulnerability Context
SNOWLOSS -.127 -.051 0.527DROUGHTLOSS -.355 -.140 0.124
Asset pentagonHuman Capital
BIRTHDAYHHH .009 .003 0.247GOVEXTVISIT .277 .110 0.186
Social CapitalOLCMEMB .177 .070 0.472F2FEXT .444 .173 0.165
Financial CapitalSAVGROUP .338 .134 0.078
Physical CapitalTRACTOR -.277 -.109 0.367RDDISTTODUKUZA .002 .001 0.729
Natural CapitalHERDSIZE .046 .018 0.000CATTLEBREED -.458 -.182 0.031WATERSOURCEDIST .009 .003 0.002
Transforming Processes and StructuresCATTLETAG .565 .216 0.070
Livelihood StrategiesCATTLEINCRANK -.187 -.074 0.003
Livelihood OutcomesREMITRANK .189 .075 0.001EXPPRICE -.000 -.000 0.280
Constant -18 .088 - 0.241Number of obs = 227Wald chi2(16) = 64.92Prob > chi2 = 0.000
Source: Authors’ survey, 2012-2013.
218 JORINE T. NDORO, MAXWELL MUDHARA AND MICHAEL CHIMONYO
participation. It also provides evidence of a mo-tivational aspect in cattle marketing in SouthAfrica. To a certain extent, the results reveal thatcattle market participation in OLM is essentiallya reactive livelihood strategy, a fallback planagainst harsh environmental and/or economicconditions.
IMPLICATIONS FOR LIVESTOCKEXTENSION
This sample evidence on the effect of vari-ous livelihood factors on cattle market partici-pation has considerable implications for the de-sign of livestock extension programs in OLM,and South Africa, in general.
Table 3: Livelihood determinants of cattle mar-ket supply volumes among smallholder farmersin OLM, Results estimated using truncated re-gression model
SLF component and variable Coefficient p>|Z| nameDependent variable: TOTSOLD
1. Vulnerability ContextSNOWLOSS 2.872 0.728DROUGHTLOSS -11.825 0.318
Asset PentagonHuman Capital
BIRTHDAYHHH -.135 0.619GOVEXTVISIT .922 0.915
Social CapitalOLCMEMB -11.722 0.197F2FEXT 21.966 0.029
Financial CapitalSAVGROUP -3.275 0.631
Physical CapitalTRACTOR 7.991 0.287RDDISTTODUKUZA
Natural CapitalHERDSIZE 1.340 0.002CATTLEBREED 7.648 0.481WATERSOURCEDIST .052 0.648
Transforming Processesand Structures
CATTLETAG -14.332 0.235Livelihood Strategies
CATTLEINCRANK 5.213 0.225 Livelihood Outcome
REMITRANK 3.147 0.298EXPPRICE -.005 0.092IMR 14.140 0.349
Constant 148.752 0.774Number of obs =111Wald chi2(17) =28.09Prob > chi2 =0.030
Source: Authors’ survey, 2012-2013. Tabl
e 4:
Cor
rela
tion
mat
rix
for
inde
pend
ent
vari
able
s us
ed i
n th
e ec
onom
etri
c m
odel
SNO
W D
ROUG
H B
IRTH
GOV
EX O
LCM
F2 SA
VG TR
ARD
D H
ERD
CAT
TLE
WAT
ER C
ATTL
CATT
REM
MEX
PPLO
SS T
LOSS
DAYH
HH TV
ISIT
EM
B F
EXT
ROU
P C
TOR
ISTT
O S
IZE
BRE
ED SO
URC
ETA
GLE
INC
IRAT
E R
ICE
DUKU
ZA E
DIST
SNO
WLO
SS1
DRO
UG
HTL
OSS
0.14
521
BIR
THD
AYH
HH
0.02
410.
0796
1G
OV
EXTV
ISIT
0.12
050.
0594
-0.0
370
1O
LCM
EMB
0.11
63-0
.014
70.
0221
0.25
551
F2FE
XT
0.06
630.
1317
0.09
03-0
.171
7-0
.122
11
SAV
GR
OU
P0.
0418
0.00
650.
0303
0.06
96-0
.080
00.
0480
1TR
ACT
OR
0.06
690.
0434
0.02
670.
0913
-0.0
325
-0.0
158
-0.0
058
1RD
DIS
TTO
DU
KU
ZA-0
.085
0-0
.196
10.
1165
0.01
470.
0715
-0.0
810
0.03
31-0
.028
11
HER
DSI
ZE0.
2181
-0.0
110
0.20
580.
1714
0.00
340.
0208
-0.0
798
0.10
48-0
.040
31
CATT
LEBR
EED
0.04
43-0
.075
3-0
.058
30.
2056
0.02
16-0
.172
60.
1313
0.12
21-0
.015
80.
0845
1W
ATER
SOU
RCE
DIS
T0.
0111
0.05
98-0
.081
40.
2038
0.09
87-0
.105
1-0
.011
10.
0153
-0.1
732
0.07
710.
1657
1CA
TTLE
TAG
0.03
620.
1190
0.02
76-0
.051
00.
0406
0.11
000.
0545
-0.0
745
-0.0
042
0.12
300.
0687
-0.0
812
1C
ATTL
EIN
CRA
NK
0.15
650.
0368
0.02
710.
0360
0.13
210.
1435
0.05
250.
0670
-0.0
896
0.17
390.
0839
-0.0
773
0.15
281
REM
ITR
AN
K-0
.009
10.
1165
-0.0
196
-0.0
799
-0.1
016
0.00
220.
0473
-0.0
646
-0.2
057
-0.1
394
-0.0
122
0.06
18-0
.069
1-0
.168
01
EXPP
RIC
E-0
.027
30.
0202
0.01
50-0
.049
4-0
.036
20.
0799
-0.0
168
0.01
41-0
.212
1-0
.004
80.
0779
0.08
96-0
.025
40.
0440
0.00
011
CATTLE COMMERCIALIZATION IN RURAL SOUTH AFRICA 219
The significantly positive effect of farmer-to-farmer extension on cattle market supply sug-gests that farmer extension groups are key play-ers in the livestock market development. Publiclivestock extension systems designed in such away to support cattle farmers’ group formationand involvement, not just as “contact groups”that transmit messages from public extensionstaff, but as active players of extension servicefunction (Anandajayasekeram et al. 2008; Swan-son and Rajalahti 2010) are therefore expectedto spur smallholder cattle farmer’s ability to par-ticipate in rural markets. This objective can beachieved through creation of an open, demo-cratic and supporting environment throughwhich these groups can thrive, supporting ca-pacity building to improve their management,injecting basic resources to improve their inter-nal functioning, and extending their links withother group (DFID 1999). This result supportsthe imminent role of a pluralistic extension envi-ronment for a pro-poor cattle market develop-ment strategy.
This latter strategy has even further appeal.Given the significantly positive effect of accessto saving groups for cattle marketing in the re-gion, the livestock extension programs that ex-tend cattle farmers’ access to financial institu-tions are expected to scale up market participa-tion among smallholders in South Africa. Thiscan be achieved by encouraging individual cat-tle farmers and their organizations to save mon-ey, with programs for tailoring inclusive finan-cial products and for overcoming barriers relat-ed to lack of collateral, for example, by identify-ing mechanisms that enable farmers’ natural cap-ital to act as collateral (DFID 1999).
The finding that endowment in natural capi-tal matters for market participation implies thatcontinued efforts by agricultural extension touplift the productivity of local breeds throughaccess to quality feeds and veterinary servicesare expected to increase both cattle productivi-ty and market participation in OLM. These ef-forts can be channeled through a centralizedcommodity-specialized approach using commod-ity extension models to ensure access to re-quired inputs including technology and financeas well as financial gains, although farmers’ feed-back can be better accounted for by other ap-proaches (Anandajayasekeram et al. 2008). How-ever, this productivity emphasis does not needto detract attention from complexity of the is-sues surrounding the management of natural
ecosystems such as water and land for livestock-based livelihoods development in the area. Thiscomplexity requires participatory and systemapproaches to agricultural extension such as thefarming system development approach built ona broader understanding of livelihood systems,the combination with other assets to sustain live-lihoods, the role of structures and processesthat govern the use of these resources (envi-ronmental laws, land and water allocation sys-tems).
The significant effect of compliance withLivestock Identification Act implies that publicextension strategy focusing on programs thataims to alleviate this challenge can be expectedto unlock markets for stallholders. These chal-lenges, as outlined in Coetzee et al. (2005) andGroenewald and Jooste (2012), include high costof registering unique brands and marking andbranding equipments, high possibility of filingclaims after stray animals cause road accident orintrude neighboring fields, and lack of brandingand marking facilities. Notably, direct supportduring the identification process such as pene-trating villages with branding and marking facil-ities or subsidizing the branding cost can beenvisaged (Groenewald and Jooste 2012). Equal-ly appropriate is an indirect support throughstructures that represent smallholder cattle farm-ers to expand their scope to include fast track-ing the identification process both in terms ofaccessibility of facilities and cost reduction forsmallholder farmers and access to appropriateforums for decisions making (DFID 1999).
On the significant effect of livelihood strate-gies, the public livestock extension service sys-tem that take into account the diversity of liveli-hood strategies when designing their programscan have a significant positive effect on marketparticipation. FEGs models can be best devel-oped around households that depend more onincomes from cattle sales, and those that haveless sources of unearned incomes (remittancesand pensions), if market participation is to bedeveloped. Once again, these results supportthe need for a participatory extension programto gauge the extent to which different cattle farm-ing systems rely on cattle incomes.
Finally, the results that price incentives drivecattle market supply volumes auger well with apositive expectation from the public extensionframework that advocates for a good function-ing of institutions that facilitate positive marketoutcomes by reducing the associated transac-
220 JORINE T. NDORO, MAXWELL MUDHARA AND MICHAEL CHIMONYO
tion risks and costs in order to sustain betterreturns to cattle farming and increase its attrac-tiveness. Validating the need for FEGs as previ-ously outlined, the provision of cattle marketinformation to potential participants throughfarmers’ social capital is an important strategyto integrate them into lucrative market chainsand high value channels. Using the mass mediaextension method to disseminate market infor-mation also remains an appropriate option. How-ever, as Coetzee et al. (2005) cautions, the formatof the information needs to be well understand-able by the target farmers (for example, farmerscannot clearly estimate the total value of theircattle based on information on beef price perkilogram live weight).
NOTE1 For example, i t is during this phase that the re-
searchers were informed about the extent to whichcattle prices fetched by smallholders varies consid-erably across communities, which triggered the needto test the factors motivating cattle marketing inOLM.
REFERENCES
Alene AD, Manyong VM, Omanya G, Mignouna HD,Bokanga M, Odhiambo G 2008. Smallholder marketparticipation under transactions costs: Maize sup-ply and fertilizer demand in Kenya. Food Policy, 33:318–328.
Anandajayasekeram A, Puskur R, Workneh S, Hoek-stra D 2008. Concepts and Practices in AgriculturalExtension in Developing Countries: A Source Book.Nairobi: International Livestock Research Institute.
Bahta ST, Bauer S 2007. Analysis of the Determinantsof Market Participation within the South AfricanSmall-scale Livestock Sector: Utilisation of Diversi-ty in Land Use Systems: Sustainable and OrganicApproaches to Meet Human Needs. Tropentag, Oc-tober 9-11, 2007, Witzenhausen.
Barrett CB 2008. Smallholder market participation:Concepts and evidence from eastern and southernAfrica. Food Policy, 33: 299–317.
Bayer W, Alcock R, Gillies P 2004. Going Backwards?-Moving Forward?-Nguni Cattle in Communal Kwa-Zulu-Natal. Rural Poverty Reduction Through Re-search for Development and Transformation. A Sci-entific Paper Presented at a Conference held at Ag-riculture and Horticultural Faculty, Humboldt-Uni-versitat zu, Berlin.
Behnke RHJ 1987. Cattle accumulation and the com-mercialization of the traditional livestock industryin Botswana. Agricultural Systems, 24(1): 1–29.
Bellemare F, Barrett CB 1993. An ordered Tobit modelof market participation: Evidence from Kenya andEthiopia. American Journal of Agricultural Eco-nomics, 88(2): 324-337.
Bollinger A 2007. Is Zero-Till an Appropriate Agricul-tural Alternative for Disadvantaged Smallholders of
South Africa? A Study of Surrogate Systems andStrategies, Smallholder Sensitivities and Soil Gly-coproteins. PhD Thesis, Unpublished. Copenhagen:University of Copenhagen.
Boughton D, Mather D, Barrett CB, Benfica R, AbdulaD, Tschirley D, Cunguara D 2007. Market participa-tion by rural households in a low-income country:An asset-based approach applied to Mozambique.Faith and Economics, 50: 64-101.
Carter MR, Lybbert TJ 2012. Consumption versus as-set smoothing: Testing the implications of povertytrap theory in Burkina Faso. Journal of Develop-ment Economics, 99: 255–264.
Coetzee L, Montshwe BD, Jooste A 2005. The market-ing of livestock on communal lands in the EasternCape Province: Constraints, challenges and impli-cations for the extension services. South AfricanJournal of Agricultural Extension, 34: 81-103.
Cragg JG 1971. Some statistical models for limited de-pendent variables with application to the demandfor durable goods. Econometrica, 39(5): 829-844.
Delgado CL, Rosegrant MW, Meijer S 2001. Livestockto 2020: The Revolution Continues. Paper present-ed at the Annual Meeting of the International Agri-cultural Trade Research Consortium (IATRC), Auck-land, New Zealand, January 18-19, 2001.
Department for International Development (DFID)1999. Sustainable Livelihoods Guidance Sheets -Framework. London: DFID.
Doran MH, Low ARC, Kemp RL 1979. Cattle as astore of wealth in Swaziland: Implications for live-stock development and overgrazing in eastern andsouthern Africa. American Journal of AgriculturalEconomics, 61: 41-47.
Duvel GG 2000. Towards an appropriate extension ap-proach for agriculture and rural development in SouthAfrica. South African Journal of Agricultural Ex-tension, 29: 10-23.
Elledoubt R 2012. Are Conservation Agriculture Prac-tices an Interesting Option for the Smallholder Farm-er Communities of the Okhahlamba Local Munici-pality, KwaZulu-Natal, South Africa? Masters The-sis, Unpublished. Gent: Universiteit Gent.
Groenewald JA, Jooste A 2012. Smallholders and live-stock. In: HD van Schalkwyk, JA Groenewald, GCGFraser, A Obi, A van Tilburg (Eds.): Unlocking Mar-kets: Lessons from South Africa. Mansholt Publica-tion Series - Volume 10. Wageningen: WageningenAcademic Publishers, pp. 113-131.
Ground M, Koch S 2008. Hurdle models of alcohol andtobacco expenditure in South African households.South African Journal of Economics, 76(1): 132-143.
Jarvis LS 1980. Cattle as a store of wealth in Swaziland:Comment. American Journal of Agricultural Eco-nomics, 62(3): 606–613.
Key N, Sadoulet E, de Janvry A 2000. Transactionscosts and agricultural household supply response.American Journal of Agricultural Economics, 82:245-259.
Kirsten JF, Doward AR, Poulton C, Vink N (Eds.) 2009.Institutional Economics Perspectives on African Ag-ricultural Development. Washington: InternationalFood Policy Research Institution.
CATTLE COMMERCIALIZATION IN RURAL SOUTH AFRICA 221
Krantz L 2001. The Sustainable Livelihood Approachto Poverty Reduction: An Introduction. Stockholm:SIDA.
Laper ML, Holloway G, Ehui S 2003. Policy optionspromoting market participation among smallholderlivestock producers: A case study from the Philip-pines. Food Policy, 28: 187-211.
Lehloenya KC, Greyling JPC, Schwalbach LMJ 2007.Small-scale livestock farmers in the peri-urban areasof Bloemfontein, South Africa. South African Jour-nal of Agricultural Extension, 36: 217-228.
Low ARC, Kemp RL, Doran MH 1980. Cattle wealthand cash needs in Swaziland: Price response and ruraldevelopment implications. Journal of AgriculturalEconomics, 31: 225-236.
Makhura MT 2001. Overcoming Transaction CostsBarriers to Market Participation of SmallholderFarmers in the Northern Province of South Africa.PhD Thesis. Pretoria: University of Pretoria.
Mathenge M, Place F, Olwande J, Mithoefer D 2010.Participation in Agricultural Markets among thePoor and Marginalized: Analysis of Factors Influ-encing Participation and Impacts on Income andPoverty in Kenya. Study Report. Tegemeo Insti-tute, Egerton University.
Matshe G, Young T 2004. Off-farm labour allocationdecisions in small-scale rural households in Zimba-bwe. Agricultural Economics, 30: 175-186.
Moffatt PG 2005. Hurdle models of loan default. TheJournal of the Operational Research Society, 56(9):1063-1071.
Montshwe BD 2006. Factors Affecting Participationin Mainstream Cattle Markets by Small-scale CattleFarmers. MSc Thesis. Bloemfontein: University ofFree State.
Musemwa L, Chagwiza C, Sikuku W, Fraser G, Chi-monyo M, Mzileni N 2007. Analysis of cattle mar-keting channels used by small scale farmers in theEastern Cape Province, South Africa. Livestock Re-search for Rural Development, 19(9): Article 3.
Musemwa L, Mushunje A, Chimonyo M, Mapiya C2010. Low cattle market off-take rates in commu-nal production systems of South Africa: Causes andmitigation strategies. Journal of Sustainable Devel-opment in Africa, 12(5): 209-226.
Musemwa L, Mushunje A, Chimonyo M, Fraser G,Mapiye C, Muchenje V 2008. Nguni cattle market-
ing constraints and opportunities in the communalareas of South Africa: Review. African Journal ofAgricultural Research, 3(4): 39-245.
National Department of Agriculture (NDA) 2006. Live-stock Development Strategy for South Africa. De-partment of Agriculture, Forestry and Fisheries, Re-public of South Africa.
National Department of Agriculture (NDA) 2010. Ag-ricultural Marketing Strategy for the Republic ofSouth Africa. Department of Agriculture, Forestryand Fisheries, Republic of South Africa.
National Department of Agriculture (NDA) 2011. AProfile of the South African Beef Market Value Chain.Department of Agriculture, Forestry and Fisheries,Republic of South Africa.
Nkosi SA, Kirsten JF 1999. The marketing of live-stock in South Africa’s developing areas: A case studyof the role of speculators, auctioneers, butcheriesand private buyers in Lebowa. Agrekon, 32(4): 230-237.
Okhahlamba Local Municipality (OLM) 2011. Okhahl-amba Local Municipality Draft IDP Review-2011/2012. Okhahlamba Local Municipality.
Okhahlamba Local Municipality (OLM) 2012. FinalIn tegrated Plan 2012-2017 . Okhahlamba LocalMunicipality.
Rios AR, Shively GE, Masters WA 2009. Farm Produc-tivity and Household Market Participation: Evidencefrom LSMS Data. Contributed Paper Prepared forPresentation at the International Association ofAgricultural Economists Conference, Beijing, Chi-na, August 16-22, 2009.
Swanson BE, Rajalahti R 2010. Strengthening Agricul-tural Extension and Advisory Systems: Proceduresfor Assessing, Transforming and Evaluating Exten-sion Systems. Agricultural and Rural DevelopmentDiscussion Paper 45.Washington: The World Bank.
Uchezuba1 UI, Moshabele E, Digopo D 2009. Logisti-cal estimation of the probability of mainstreammarket participation among small-scale livestockfarmers: A case study of the Northern Cape prov-ince. Agrekon, 48(2): 171-183.
Winter-Nelson A, Temu A 2005. Impacts of prices andtransactions costs on input usage in a liberalizingeconomy: Evidence from Tanzanian coffee grow-ers. Agricultural Economics, 33: 243–253.
Wooldridge JM 2002. Econometric Analysis of CrossSection and Panel Data. Cambridge, MA: MIT Press.
top related