Factors influencing participation in dairy goat milk marketing in Kenya and its implication for a sustainable breeding program

T D O Ogola and I S Kosgey^1,2

Department of Agricultural Economics and Agribusiness Management, Egerton University, P O Box 536-20115 Egerton, Kenya
isaacs.kosgey@gmail.com
¹ Department of Animal Sciences, Egerton University, P O Box 536-20115 Egerton, Kenya
² Moi University, P O Box 3900-30100 Eldoret, Kenya

Abstract

Dairy goat production is multi-faceted and various organizations in Kenya have promoted it with the main objective being to make families food secure. However, marketing plays an important function in ensuring better income that would improve the welfare levels of dairy goat farmers. Dairy goat milk marketing in Kenya can be described as a relatively traditional or informal affair. Organized production and processing is limited in both absolute and utilized capacity, if not non-existent, despite attempts to promote dairy goat production in the country. Lack of market infrastructure and institutions in rural areas suggest markets are thin and imperfect, leaving farmers to make their own efforts to market their milk. This has effects on sustainability of a dairy goat breeding programme. The current study examined factors that contributed to market participation of dairy goat farmers in Kenya and the implication for a sustainable breeding programme. With the use of Heckman estimation procedure, the study identified policy and technology options to increase participation and sale of milk by dairy goat farmers. Data used was collected from 71 household in a cross-sectional survey conducted in the Rift Valley, Nyanza and Coast regions of the country. Results indicate that physical capital (i.e., size of land and livestock owned), milk price, distance to the market and age of the household head influenced market participation and dairy goat milk sales. Dairy goat farmers in their management need to balance flock numbers with feed availability and genetic improvement. Consequently, strategies targeting optimizing available resources by matching livestock production systems to available resources, improving existing technologies and integrating technologies that use multipurpose animals and crops, and recycling of crop residues and by-products as feed for dairy goats are options for ensuring adequate nutrition for dairy goat productivity, increasing flock size and exploiting market opportunities. A balanced breeding programme combining milk yield, fertility and longevity traits in the breeding goal should be adopted. The foregoing, if addressed adequately, would lead to a sustainable dairy goat breeding programme.

Key words: breeding programme, market participation, dairy goats, Kenya, smallholder farmers

Introduction

Dairy goat farming in Kenya is an enterprise that has grown in popularity over the past decades. The option to use exotic dairy goats in the country has gathered momentum from the 1990’s to the present, mainly due to diminishing land sizes and population pressure (Ahuya et al 2005; Ogola et al 2010). Dairy goat production has multiple dimensions and objectives, with the main premise being that they quickly increase animal production and, subsequently, improve economic returns and diet quality of smallholder households adopting them (Kosgey et al 2006; Ogola et al 2009). Other studies have shown that smallholder farmers face two decision problems after production; the first being whether to sell or not to sell their produce and the second being how much to sell into a market (e.g. Goetz 1992; Heltberg and Tarp 2002; Boughton et al 2007).

Market participation has been noted to hold considerable potential for unlocking suitable opportunities necessary for providing better incomes and sustainable livelihoods for smallholder farmers (World Bank 2008; Omiti et al 2009). As the marketed share of agricultural output increases, input utilization decisions and output combinations are progressively guided by profit maximization objectives (Omiti et al 2009). Tsourgiannis et al (2010) suggest that, to increase farm profitability and sustainability of a livestock enterprise within an intensively competitive environment, farmers should focus on a market orientation strategy.

Dairy goat milk marketing in Kenya can be described as a relatively traditional or informal affair although the literature provides that improving access to markets has numerous benefits. Organized production and processing is limited in both absolute and utilized capacity if not non-existent despite a whole series of attempts to promote dairy goat production. Lack of market infrastructure and institutions in rural areas suggest that many markets are thin and imperfect. Despite existence of a number of market outlets, they are scattered and mainly found in urban areas. In view of the absence of most modern dairy goat milk plants, it seems likely that traditional production systems will continue to be important in the foreseeable future. Individually, there is lack of sufficient volumes of uniform goat milk to attract buyers to the farms, leaving farmers to make their efforts to market the milk. Most products from dairy goat milk end up being consumed within the producing household.

Various studies have identified factors considered as restraining the ability of small-scale farmers to markets as lack of market information (Goetz 1992; Omamo 1998) and large distances to the market place (Key et al 2000; Makhura et al 2001). According to Makhura et al (2001), small-scale farmers contribute inadequately to the mainstream market because of low production and poor access to other options for obtaining a livelihood. Other factors include, farmer training, land ownership, flock size, household demographic characteristics and support services (Lapar et al 2003; Bellemare and Barrett 2006; Kuma et al 2013). Barret (2008) identified pricing as a major factor in market participation.

Despite the increasing frequency of debate based on the belief about the relative benefits of dairy goat milk or on consumer willingness to pay for it, factors that influence participation and actual levels of participation by farmers in the sale of dairy goat milk are unknown. The sale of milk and milk products by the producers is important to the monetization of traditional livestock economies. Given the importance of marketing in the transition of small-scale dairy goat farmers to commercial farming, an understanding of the variables that explain smallholder farmer response to market participation is important and can help refine and better target policies. It is important to determine these variables in view of the government’s and other development agencies’ interest in encouraging domestic dairy goat production. This requires an empirical study to investigate factors affecting decisions on dairy goat milk sales and access. There is, however, a paucity of information on dairy goat milk marketing because regulation and focus in the mainstream fluid milk market in the country has largely been on dairy cow milk. The purpose of this study was, therefore, to identify factors that influence participation by farmers in dairy goat milk marketing and extent of market participation in Kenya. The implication of these on a sustainable dairy goat breeding programme is explored in the current study.

Theoretical framework

Producers are assumed to adopt an optimization behaviour considering a choice between primary marketing strategies. This approach of profit maximization assumes perfect competition. It has, however, been criticized by economists because knowledge about the future is uncertain and marketing decisions take place in an environment of uncertainty. Game and decision theories were developed to explain decision making under uncertainty. These theories must, however, be combined with that of utility analysis as described by Von Neumann and Morgenstern (2007). The utility analysis and decision theory form the household objectives approach. The producers’ best plan of action would be to choose a marketing strategy that maximizes their expected utility. It is assumed that dairy milk producers select their market channels and supplies based on their expected utility. Standard utility maximization theory stipulates that, for any given market price, each participant will derive a specific level of utility (i.e. the indirect utility).

The decision to participate in marketing is a decision related to the level of complexity of the household. This requires household models rather than a competitive market framework. This is because households make joint decisions with respect to consumption, production and labour. These models, therefore, emanate from non-separability rather than separability. Since expected utility is assumed, the optimal strategy is also conditioned on the degree of risk aversion. Obviously, an increase in marketing cost would reduce expected utility for that specific marketing strategy. The household model incorporates transaction cost in line with the procedures of Key et al (2000). However, from empirical studies, transaction costs alone are not sufficient in determining the extent of coordination in a supply chain because other factors also have a role to play (Ferto and Szabo 2002; Woldie and Nuppenau 2009; Zivenge and Karavina 2012).

Market participation can be modeled on the basis of a household or an individual facing levels of utilities, U₁ or U₀, from making choices between participation and non-participation, respectively. However, the observed state only reveals which choice provides a higher utility but not the unobservable utility. That is, the observed (latent) indicator equals 1 if U₁ > U₀ and 0 if U ₁ < U₀. By assuming that differences across utilities are determined by household or individual specific characteristics, then the decision to maximize utility is defined over consumption of a vector of agricultural commodities, Y^c for c=1.....c, and a Hicksian composite of other tradables, X. The household earns income from production and sale of any or all of the milk, M, and from off-farm earning, W. Milk is produced using technology, fm (Am, G), that maps flow of services provided by privately held assets, among them being land, labour, livestock and machinery, reflected in the vector A, and public goods and services like roads and extension services, represented by vector G, into output. The household faces a parametric market price for milk, Pm, and a vector of milk- and-household-specific transactions costs per unit that depend on public goods and services, G (.radio broadcast of prices that affects search costs and road accessibility to market), household-specific characteristics (i.e. educational attainment, gender and age) that may affect, for example, search costs and negotiation skills, reflected in the vector Z as well as the household’s assets, A, liquidity from non-farm earnings, W, and net sales volumes. The households’ choice can be represented as a constrained optimization problem where it maximizes utility subject to the cash budget, and available non-tradable resources. The model can be estimated by regressing an indicator variable Y = 1 for market participation and Y = 0 for the non-market participation on X, which is a matrix of household- or individual-specific covariates like education, skills and gender, and product attributes like quality to obtain β. The vector β is composed of unknown coefficients controlling the relationship between household- or individual-specific characteristics and market participation plus a random error.

Conceptual framework

Figure 1 below depicts the conceptual framework model for the study. The independent variables are household institutional factors, asset endowment, labour structure, market access labour and milk price, while the dependent variables are market participation and volume of milk marketed.

Figure 1. Conceptual framework model of the determinants of household level market participation

Research methodology

The study area, and data sources and type

Primary data from a cross-sectional survey of 71 smallholder farmers was used. The study was conducted in the Coast (Kwale County; lower agro-ecological zones 2 and 3), Nyanza (Homabay, Nyakach, Nyanza, Rongo, Siaya and Suba districts; low-medium potential agro-ecological zones 1 and 2) and the Rift Valley (Bomet district; lower agro-ecological zone 2) regions of Kenya, where farmers were randomly selected from a sampling frame obtained from the Heifer Project International-Kenya offices in the respective region based on having reared the goats for at least one lactation period. Data was captured through personal interviews with the use of a structured questionnaire administered in the sampled farms.

Data analysis

The data was subjected to both descriptive and inferential statistical analyses such as percentages, means, standard deviations, t-test as well as χ². The Statistical Package for Social Scientists (SPSSv17) software was used for this purpose. STATA software was used for the econometric analyses to capture factors affecting participation in marketing and intensity of participation.

Determinants of market participation and level of participation: Heckman selection model

The Heckman selection model (Heckman 1979; Greene 2003; Wooldridge 2013) assumes that there exists an underlying regression relationship as shown in equation 1 below:

y_j = x_jβ + u_1jregression equation (1)

The dependent variable, however, is not always observed. Rather, the dependent variable for observation j is observed if (equation 2 below):

z_j + u_2j > 0 selection equation Probit (2)

Where u₁ = N (0; δ ), u₂ = N (0; 1) and corr(u₁; u₂) = ƿ. When ƿ ≠ 0, standard regression techniques applied to the first equation yield biased results. Heckman selection model provides consistent, asymptotically efficient estimates for all the parameters in such models.

Results and discussion

Table 2 below depicts the characteristics of both participants and non-participants in dairy goat milk marketing. About 73.0% of the respondents involved in dairy goat farming were females while a few were men. The male to female ratio across the participant or non-participant levels were similar. The average family size was 7±4.6, with farmers selling dairy goat milk having a slightly higher trend of average family size of 7±5.9 than those not selling (6±=2.4). Increased number of family members can provide more labour that could be utilized for dairy goat production and lead to surplus. Simultaneously, their consumption needs could also be higher, which may dampen sale of milk. The average age of the dairy goat farmers was 49 years. Youth representation was overtly low, and this may raise concerns about succession in dairy goat production in the surveyed areas. Generally, those selling dairy goat milk were younger (45±0.43 years) than non-participants (53±22.19 years). Overall, the number of years spent in school was, on average, 6.28±4.18. This represented a central tendency towards primary school level of education. In terms of attainment of education, participants had spent, on average, slightly more years in school (6±4.26) than non-participants (5±4.09). This difference was not statistically significant. Averagely, land owned by farmers was 1.3±3.6 acres. The average number of dairy goats kept by the farmers was 3.1±2.3. Comparatively, the average flock sizes for those participating in the sale of dairy goat milk were higher (9.49±4.56) than for those who did not sell (2.6±1.3). The higher values for productive assets may allude to higher production potential of participants to non-participants with respect to milk production.

Factors influencing dairy milk market participation

Results of the first-stage probit model estimation of the factors that determined the decision by a smallholder dairy goat farmer to participate in a channel are presented in Table 3 below along with values of the marginal effects. The model χ² applying appropriate degrees of freedom indicate that the overall goodness of fit of the probit model was significant (P<0.01). This shows that, jointly, the model fits better than a model with no predictors. The McFadden’s Pseudo R² value obtained indicate that the independent variables included in the regression explained 84.3% proportion of the variations in the dairy goat farmers’ participation decisions. Decisions on participating in the market were positively and significantly associated with market milk price.

Market price played an important role in market participation. The marginal effect indicated that a unit increase in market price would lead to an increase in market participation by 1.1%. As opined by Randela et al (2008), low prices are a disincentive to market participation, and remote locations of farms coupled with poor road infrastructure, may contribute to high transaction costs. This can contribute to non-participation in marketing. In the current study, it is likely that high milk prices negated this effect, which encouraged market participation. According to Burke et al (2015), higher milk prices are associated with a higher probability of a producer being a net seller.

The dependent variable is a dummy variable that takes on the value 1 if the farmer participates in marketing/ selling goat milk, 0 otherwise. Figures in parentheses are robust standard errors.

Factors influencing quantity of milk sold

The factors that influenced intensity of participation in dairy goat milk marketing were estimated using a linear regression with identified variables from the Probit model while incorporating the inverse Mills ratio (λ). Table 4 below indicates the results of the second stage of the Heckman model. The λ had an insignificant effect on quantity marketed, suggesting absence of selection bias. This means that the error term of the participation in the informal channel and intensity were not correlated (Woodridge 2013). If there is no evidence of sample selection, consistency of OLS (Ordinary Least Squares) will not be affected (Wooldridge 2013), and the models return an adjusted R² of 0.85.

There was a negative relationship between age and the amount of milk sold. An increase in the age of the household head by one year decreased the probability of participating in the milk market by 20%. Barret et al (2007) observed that younger people participated more in the market because they were more receptive to new ideas and were less risk averse than the older people. Another possible explanation could be that farmers consumed more goat milk due to its therapeutic properties and, thereby, reduced the amount available for sale with age.

The quantity of milk sold was positively associated with land size. This observation concurs with that of Randela et al (2008) that access to land, measured by the size of the arable land the household operates, was a necessary condition for market participation. The larger the size of arable land a household uses, the higher the production levels are likely to be, and the higher the probability of market participation. An increase in the land size allocated towards dairy goat production may enhance the possibility of selling more milk through fodder output or increased stocking rate capacity. This is in agreement with the findings of Boughton et al (2007) to the effect that this could be due to increased quantity of output associated with greater farmer productive assets.

There was a positive relationship between market participation and distance to the market. This is in contrast to other studies that have established that increasing market distance was negatively associated with the amount of milk that was sold, i.e., there was lower marketable surplus with increasing market distance (Holloway et al 2000; Ehui et al 2003; Bardhan et al 2012). According to Ehui et al (2003), areas with greater population pressure were associated with greater sales of dairy products, indicating profitability of dairy production in densely populated areas. That is, areas with lower population pressure (e.g. rural areas and farms) are associated with lower sales of milk product, but greater sales of dairy products took place in centres with slightly greater population (e.g. urban centres).

Implications for a breeding programme

Summary statistics revealed interesting and important differences between participating and non-participating farmers with respect to dairy goat milk market participation. Generally, farmers selling dairy goat milk were relatively young. Averagely, farmers owned larger flock sizes, which coincided with larger land acreage, suggesting availability of pasture and other feeds for the dairy goats. To ensure farmers exploit market opportunities, feed availability and management must be balanced with flock numbers and genetic improvement of the dairy goats. Econometric results indicated that participation in dairy goat milk marketing was likely to be influenced by milk price and available land resources. There was a positive relationship between market participation and distance to the market. Distance is indicative of lack of market nearby or existence of a specialized market.

Components of strategies to ensure adequate nutrition could entail selection of crops and cropping systems that will maximize biomass production and nitrogen fixation, with minimal use of inputs external to the system. Alternatively, the strategy can involve planting high yielding fodder or pasture, use of fertilizer to improve pasture yields, integration of dairy goats in crop production, increased use of legumes in cropping, and more effective use of crop residues and agro-industrial by-products. This must be complemented by conservation of forage as silage for feeding dairy goats during the dry season. Increased conservation of feed as cut or standing hay and use of supplements would also boost production of dairy goats.

It was evident that participation in dairy goat milk marketing was positively correlated to owning more goats. To improve on milk production alone may involve development of productive dairy goat lines that are predictable in performance, which entails commitment to line breeding. The best programme in this situation is one that will combine production, fertility and longevity traits in a total merit index. That is, line breeding for bucks to improve on milk traits but outbreeding for does. This may necessitate enlisting of progressive farmers in a group selection scheme, given the existing challenges in obtaining dairy goats from institutions and or having specialized breeding farms. Since only natural mating took place amongst the flocks, selection from various village dairy goats will involve judgement on milking ability, size, conformation and prolificacy. It is important to ensure that bucks are selected on doe records and be replaced annually to minimize inbreeding. Keeping and maintenance of records by the farmers is, therefore, a prerequisite to ensure the success and sustainability of such a programme. Improving the overall environment in which the dairy goat farmers operate like improving basic governance systems, policy, infrastructure, technology, education and institutional capacity could significantly assist in addressing challenges in market participation. This must be undertaken while making the right selection of dairy goats and taking into consideration the farmers' overall breeding objective. The breed of choice given to farmers should be on the basis of participatory assessment of the production system, taking into consideration the farmers’ production aspirations.

Acknowledgements

The authors are grateful to Egerton University, Moi University and Heifer Project International, Kenya (HPIK) for the support for the study.

References

Ahuya C O, Okeyo, A M and Murithi F M 2005 Productivity of cross-bred goats under smallholder production systems in the Eastern highlands of Kenya 1. Small Stock in Development 54.

Bardhana D, Sharma M L and Saxena R 2012 Market participation behaviour of smallholder dairy farmers in Uttarakhand: a disaggregated analysis. Agricultural Economics Research Review 25(2): 243-254.

Barrett C B 2008 Smallholder market participation: concepts and evidence from Eastern and Southern Africa. Food Policy 33(4): 299-317.

Bellemare M F and Barrett C B 2006 An ordered Tobit model of market participation: evidence from Kenya and Ethiopia. American Journal of Agricultural Economics 88: 324-337.

Boughton D, Mather D, Barrett C B, Benfica R, Abdula D, Tschirley D and Cunguara B 2007 Market participation by rural households in a low-income country: an asset-based approach applied to Mozambique. Faith and Economics 50(1): 64-101.

Burke W, Myers R and Jayne T S 2015 A triple-hurdle model of production and market participation in Kenya’s dairy market. American Journal of Agricultural Economics 97(4): 1227–1246.

Ehui S, Benin S and Paulos Z 2009 Policy Options for Improving Market Participation and Sales of Smallholder Livestock Producers: A Case Study of Ethiopia. Draft Prepared for Presentation at the 27th Conference of the International Association of Agricultural Economists (IAAE), 16-22 August 2009, Beijing, China.

Ferto I and Szabo G G 2002 The choice of supply channels in Hungarian fruit and vegetable sector. In: Proceedings of the American Agricultural Economics Association Annual Meeting, July 28-31, 2002, Long Beach, CA, USA.

Goetz S 1992 A selectivity model of household food marketing behavior in Sub-Saharan Africa. American Journal of Agricultural Economics 74: 444–52.

Greene W H 2003 Econometric Analysis, 5th Edition, Upper Saddle River, New Jersey, USA: Pearson Education, Inc.

Heckman J J 1979 Sample selection bias as a specification error. Econometrica 47: 153-161.

Heltberg R and Tarp F 2002 Agricultural supply response and poverty in Mozambique. Food Policy 27: 103–124.

Holloway G, Nicholson C, Delgado C, Staal S and Ehui S 2000 Agroindustrialization through institutional innovation transaction costs, cooperatives and milk-market development in the East-African highlands. Agricultural Economics 23: 279-288.

Key N, Sadoulet E, and de Janvry A 2000 Transaction costs and agricultural household supply response. American Journal of Agricultural Economics 82(1): 245–59.

Kosgey I S, Baker R L, Udo H M J and van Arendonk J A M 2006 Successes and failures of small ruminant breeding programmes in the tropics: a review: Small Ruminant Research 61: 13-28.

Kuma B, Baker D, Getnet K and Kassa B 2013 Factors affecting milk market participation and volume of supply in Ethiopia. Asian Journal of Rural Development 4(1): 1-15.

Lapar M L, Holloway G and Ehui S 2003 Policy options promoting market participation among smallholder livestock producers: a case study from the Phillipines. Food Policy 28(3): 187-211.

Makhura M, Kirsten J and Delgado C 2001 Transaction costs and small holder participation in the maize market in the Northern Province of South Africa. In: Seventh Eastern and Southern African Regional Maize Conference, pp. 463-462.

Ogola T D O, Nguyo W K and Kosgey I S 2009 Economic contribution and viability of dairy goats in Kenya: implications for a breeding programme. Tropical Animal Health and Production 42: 875-885.

Ogola T D O, Nguyo W K and Kosgey I S 2010 Dairy goat production practices in Kenya: implications for a breeding programme. Livestock Research for Rural Development 22(1), Article #16.

Omamo S W 1998 Farm-to-market transaction costs and smallholder agriculture: explorations with a non-separable household model. Journal of Development Studies 35: 152-63.

Omiti J, Otieno D, Nyanamba T and McCullough E 2009 Factors influencing the intensity of market participation by smallholder farmers: A case study of rural and peri-urban areas of Kenya. African Journal of Agricultural and Resource Economics 3(1): 57-82.

Randela R, Alemu Z G and Groenewald J A 2008 Factors enhancing market participation by small-scale cotton farmers. Agrekon 47(4): 451-469.

Rios A R, Shively G E and Masters W A 2009 Farm productivity and household market participation: evidence from LSMS data. In: Proceedings of the International Association of Agricultural Economists Conference, August, 2009, Beijing, The Peoples’ Republic of China.

Tsourgiannis L, Karasavvoglou A, Eddison J and Warren M 2010 Profiles of dairy cow farmers’ marketing strategies in the county of Cornwall in the UK. Intellectual Economics 2(8): 57–73.

Von Neumann J and Morgenstern O 2007 Theory of games and economic behavior. Princeton University Press, 46 pp.

Woldie G A and Nuppenau E A 2009 Channel choice decision in the Ethiopian banana markets: a transaction cost economics perspective. Journal of Economic Theory 3(4): 80-90.

Wooldrige J M 2013 Introductory Econometrics A Modern Approach. 5th Edition, Natorp Boulevard Mason, OH 45040 USA: South-Western 5191, 910 pp.

World Bank 2008 World Development Report 2008: Agriculture for Development, Washington D.C., USA.

Zivenge E and Karavina C 2012 Analysis of factors influencing market channel access by communal horticulture farmers in Chinamora District, Zimbabwe. Journal of Development and Agricultural Economics 4(6): 147-150.

Received 24 February 2019; Accepted 24 February 2019; Published 1 April 2019

Go to top