Previous Page Table of Contents Next Page

Constraint identification and analysis in African small ruminant systems - Identification et analyse des contraintes dans les systèmes de production des petits ruminants en Afrique

P.K. Ngategize

International Livestock Centre for Africa
P.O. Box 5689
Addis Ababa

Small ruminant production and economic roles
Economics of small ruminant production
Models and approaches to constraint identification and analysis
Use of simulation for constraint identification and analysis


A review of the economic role of small ruminants and their production constraints is presented. Approaches to small ruminant constraint identification and analysis are identified. Levels of production economics analysis are suggested and examples of economic or statistical models given. Simulation analysis is identified as a powerful tool in constraint identification and analysis due to its robustness.


Le rôle économique et les contraintes de production des petite ruminants vent passés en revue. Quelques approches à l'identification et l'analyse des contraintes vent décrites. Des niveaux d'analyse vent suggérés et des exemples des modèles économique et statistiques vent donnés. Un modèle de simulation est identifié comme un outil puissant pour l'identification et l'analyse des contraintes


Only during the last decade has smallholder livestock production in sub-Saharan Africa received the attention of economists. Previous interest was limited to commercial production and the major pastoral and agropastoral systems. In the former, research focussed on breeding, disease control and other aspects based on western systems, while in the latter research was mainly dominated by anthropologists and sociologists. Livestock in smallholder systems were considered as scavengers, surviving on crop residues and household wastes (Deans, 1981). The economic importance of small ruminants in mixed systems now receives more attention, with increases in resource allocations to research by many national and international institutions.

Much of current small ruminant research, nonetheless, is dominated by descriptions of production systems and traits (ILCA, 1979; Gatenby and Trail, 1982; Wilson and Bourzat, 1985; Sumberg and Cassaday, 1986). Little economic analysis of the frequently reported constraints has been done. The economic role of goats and sheep has, however, been described and some economic analysis of technological innovations and production prospects has been conducted (Upton, 1984; 1985; Gryseels et al, 1986; Sumberg et al, 1987).

Economic analysis of alternative management techniques and evaluation of the socio-economic feasibility of potential technological innovations must go hand in hand. If this is not done it may be discovered too late that efforts to improve small ruminant production are destined to fail or will produce limited impact as has been the case with investments in pastoral production systems. The objectives of this paper are to:

review the literature on small ruminant production economics in sub-Saharan Africa;

identify existing or potential analytical approaches to the economics of production; and

assess the potential role of simulation in constraint identification and analysis.

Small ruminant production and economic roles

In smallholder production systems, goats and sheep are important because they require low initial capital and maintenance costs, are able to use marginal land and crop residues, produce milk and meat in readily usable quantities, and are easily cared for by most family members. Small ruminants are prolific and need only short periods to increase flock sizes after catastrophies or in periods of high prices and thus offtake rate can respond to price increases (Winrock International, 1983). The sheep enterprise in the Ethiopian highland crop and livestock system is the most important form of investment and cash income and provides social security in bad crop years (Getachew, 1988). In Yatenga, Burkina Faso, livestock cash income was 33% to 99% of total farm cash income with goat and sheep sales providing 52% of livestock cash income (Bourzat, 1985).

Given these advantages, it is not surprising that goats and sheep are found in many smallholder systems. Growth rates have, however, been low at about 1.2% per annum. World Bank livestock projects have generally produced less than 10% rates of return on investment, which is low compared to crop production (ILCA, 1987). Although some analyses (Upton, 1984) show high rates of return from small ruminants (35% for goats and 55% for sheep), this may not be realistic given the assumptions that there are no opportunity costs of labour and land. The argument that production may be limited by the relatively high risk and high initial investment costs has been advanced elsewhere (ILCA, 1979).

At the macro level, a number of production constraints have been identified: natural disasters (drought in mid-1970s and mid-1980s and floods); national policies on pricing; credit; land tenure; extension services and research; and over-valued exchange rates which encourage imports and discourage exports. International constraints include trade barriers and in some countries the surpluses that encourage "dumping" in Africa. Smallholder producers have no control over these problems and their major concerns are related to nutrition, disease, management, breeding and marketing. In Nigeria farmers ranked constraints as feed, need of fencing, time, cash (capital) and disease (Okali, 1979).

Economics of small ruminant production

The economics discipline has a broad mandate. Farmers' goals and objectives - what the farmer attempts to maximise or minimise in his production activities - have first to be identified.

Production constraints and resource limitations then need to be identified and quantified. Given the physical and sociopolitical environment, technological innovations have to be tested for adoption. In a step by step process, economics needs to provide answers to whether small ruminants are profitable as an enterprise, whether they are profitable or competitive relative to existing farm and non-farm enterprises or opportunities, what are the losses (cash or kind) associated with a given constraint to the farmer or the production system, what are the losses and benefits of introducing a given technology, and what is the optimal level of resource or technological innovation for alternative resource levels or management strategies.

Apparent production constraints such as "high" mortality, "long" birth intervals and "slow" growth rates may not be as critical to the farmer as production scientists think. Recommended technologies may not therefore be adopted in given social, economic and ecological circumstances. High mortality rates in pastoral systems may, reflect the management system - lack of permanent settlement and hence lack of housing and attention to the young. Although monetary losses may appear to be high and static monetary benefits to outweigh the costs of interventions, such interventions would impose on the herdsman the need to settle at one place to the detriment of the mature and productive livestock. Similarly, increased incomes as a result of an innovation may lead to change in farmers' attitudes towards livestock.

The question as to what type of analysis is appropriate thus arises at the individual technology or at the production system level. There is also a need to differentiate between a partial analysis of one period and a dynamic long-term analysis. The choice of model depends on the nature of the problem and the intended use of the results. This, in turn, requires the analyst to be conversant with the system being analysed. The most frequent recommendation to resolve these complex problems has been to use multidisciplinary teams. If this is not possible, it would be recommended that each researcher has to have a multidisciplinary training and approach to minimize disciplinary biases.

Models and approaches to constraint identification and analysis

Constraint identification and analysis is concerned with identifying, quantifying and conducting sensitivity analysis on technical and economic factors that may limit the profitability and hence future prospects of small ruminant production. A number of models have been developed by various disciplines. With minor modifications the models have been adopted by other disciplines. Various statistical/ epidemiological and economic models have been used in the evaluation of the impact of disease on animal production and analysis of alternative health management strategies (Ngategize and Kaneene, 1985).

The statistical/epidemiological models include regression, path and discriminant analysis and are essentially used to determine relationships and their magnitude or significance, and in disaggregating impacts. The economic models use technical data (usually generated by statistical and epidemiological models) with economic data to measure the monetary impacts and evaluate possible interventions. These models include partial budgeting, cost benefit analysis, decision analysis, linear programming, Markov processes, systems simulation and dynamic programming. Variants of linear programming have been developed including integer programming and N-stage programming. Some models can be considered as sub-components of others as, for example, when cost-benefit analysis is done within a dynamic programming framework. The choice of model depends on the experience of the user, the nature of the problem, the importance of the results in decision making, and the availability of computer software. Three analytical models - partial budgeting, cost-benefit analysis and linear programming - have been identified for the economic analysis of animal disease control.

Some of these models have already been used in the economic analysis of interventions. Partial budgeting has shown that a health package and supplementary feeding with Leucaena foliage may be profitable in the humid zone (Sumberg et al, 1987). Linear programming has been used in evaluating the economics of alley cropping (Raintree and Turay, 1980) and leguminous tree crops in zero-tillage cropping systems (Verinumbe et al, 1984). Since goat and sheep production is an enterprise in smallholder agriculture and competes for resources with other farm and non-farm enterprises, linear programming can be a powerful tool in assessing whether small ruminant production is economical within a whole farm perspective, at what level (number and resource requirement) the small ruminant enterprise should be operated (if it exists in the optimal combination), the range of prices (input/output) over which the enterprise would be economically feasible and its riskiness (incorporated through a stochastic model). Linear programming can therefore be used not only in constraint identification and analysis but also in assessing production prospects at different levels of operation (Table 1).

One limitation of most models is that they are static or steady state, with the analysis covering one period. They do not capture the dynamic nature of the problem such as the technological adaption process or the time prior to reaching the steady state. To incorporate more than one period in the analysis it has to be repeated for each period (by using a multi-period linear programming model for example). This makes the process complex, especially if, in addition, data are not available for each analysis period. Some of these problems can be overcome by using simulation analysis.

Use of simulation for constraint identification and analysis

Simulation attempts to emulate real life conditions using simple models. This is usually achieved by developing a model of variable interrelationships (usually in the form of modules) and running them in a sequence of steps over time. These models generate data where actual time series data are not available or would have been too expensive (in money and time) to generate through experiments or through the long term observation of technologies or management strategies (de Boer et al, 1982; Blackburn et al, 1984; Upton, 1986). In addition, the computation of economic outputs such as net present values and internal rates of return is made simple, especially where complex computations have to be made over many periods.

Table 1. Levels and models of production economics analysis of small ruminants.


Problem/ Question

Examples of economic models


Small ruminant enterprise profitability

Partial budgeting
Cost-benefit analysis


Relative profitability or competitiveness of the small ruminant entreprise

Gross margin analysis
Linear programming


Economic feasibility of production interventions and programmes

Cost-benefit analysis
Partial budgeting
Decision analysis


Small ruminant production, supply and demand prospects

Econometric model


Socio-cultural factors and policy effects on welfare and equity issues in small ruminant production

Socioeconomic, descriptive and perspective studies

A number of bioeconomic simulation models have been developed and applied in the analysis of beef production systems (Chudleigh and Cezar, 1982). In most cases, the objectives of such models include the simulation of physical and financial functions, assessment of the economics of various systems and available technologies, simulation of production under a range of management in various environments, investigation and demonstration of the economic effects of varying the parameters (herd size, type of animal, price, availability of credit), and suggestions for future areas of research and provision of data for extension.

To meet these objectives, a simulation model can be useful in identifying resource or production constraints within a system by changing the levels of resource use or availability in the model and through sensitivity analysis. In small ruminant production the major components of economic interest include nutrition or feed/pasture supplementation, reproduction and production efficiency (e.g. birth rates, mortality, age at first birth and growth rates), decision management variables relating to selling or purchasing and risk management, and micro- and macro-economic variables including price, market structure, credit and other policies. These production constraints can be incorporated and analysed as independent modules by changing the parameter levels and simulating their impact over time. The technical results of the simulation can then be used internally (within the model) or externally to compute economic decision parameters such as the net present value and internal rate of return.

A simulation model is useful in analysing specific or individual production constraints or technological innovations and analysing more than one constraint or innovation simultaneously. In either case, the analysis may be done at the individual animal, the flock or the systems level. Model validation can be done in relation to existing experimental and production systems results and also in relation to other simulation models (Blackburn et al, 1984). Documentation and transferability of any model are important evaluation criteria. With simulation, the economist may actively define data needs for improved analysis and hence influence experimental designs by technical scientists rather than being a mere user of the results generated.

Simulations have been performed to determine genetic potentials for size (growth rate), maturing rate and milk production for dual purpose goats in western Kenya (Cartwright et al, 1986). They showed that the environment was capable of sustaining non-indigenous breeds which could produce more milk and heavier kids. Variations in forage quality and quantity were the main factors determining the number of goats a farm could support, a flock size of 6 does being required to provide milk for an average family. The findings are said to have implications for improvements in genetic potential and availability of alternative feeds.

Steady state flock structures and annual offtakes have been used to determine internal rates of return to goat production in the humid zone of Nigeria (Upton, 1985). Sensitivity analysis showed that variations in mortality had the greatest impact on economic performance, followed by reproductive performance. Variations in price had the least effect. Such results are useful in defining priority areas for improving overall returns.

At the farm level, there is still limited knowledge of the input-output coefficients that are important in economic analysis. Most inputs and outputs are not marketed and it is difficult to put monetary values on them. It is recognized that smallholder management incorporates risk and uncertainties in decisions which are difficult to measure and vary in type and in magnitude from one production system to another. Socio-economic factors (child age, education, polygamy, cultural attitudes) also have an impact on production but are also difficult to quantify and incorporate in economic models. Such problems may be highlighted in the discussion of results or, if necessary, be analysed or ranked within the expected utility hypothesis framework.


The importance of small ruminants in smallholder production in Africa has been well documented. On- and off-farm production constraints have been identified in broad terms. Technological innovations have been suggested but the evaluation of their economic significance, within the various production systems, has not yet been sufficiently analysed. Economic analysis of small ruminant production can be conducted at several levels which depend mainly on the objectives of the analysis and on data availability. Partial or static models are not sufficient in a systems approach to analysis. Computerized simulation analysis provides a powerful tool for identifying and analysing production constraints and evaluating the economic impact of technological innovations and alternative management strategies.


Comments on earlier drafts of this paper were made by R Brokken, R T Wilson, K J Peters and Getachew Assamenew.


Blackburn H D, Cartwright T C, Carles A B and Ruvuna F. 1984. A conceptual overview of the SR-CRSP goat production model. Proceedings of the 3rd SR-CRSP Kenya Workshop. Small Ruminant Collaborative Research Support Programme, Nairobi, Kenya.

Bourzat D. 1985. The role of small ruminants in the agropastoral farming systems in Yatenga province (Burkina Faso). In: I F Adu, O A Osinowo, B B A Taiwo and W S Alhassan (eds.), Small ruminant production in Nigeria. National Animal Production Research Institute, Shika-Zaria, Nigeria.

Cartwright T C, Blackburn H D and Ruvuma F. 1986. Development of dual purpose goat production in Kenya - A systems analysis paradigm. Proceedings of the 5th SR-CRSP Kenya Workshop. Small Ruminant Collaborative Research Support Programme, Nairobi, Kenya.

Chudleigh P C and Cezar I M. 1982. A review of bioeconomic simulation models of beef production systems and suggestions for methodological development. Agricultural Systems 8: 273-289.

Deans R. 1981. Farming systems research as it relates to the animal sciences. Farming Systems Research Group Working Paper No. 5. Michigan State University, East Lansing, USA.

de Boer A J. Job M and Matthewman R W. 1982. Economic modelling of small farms in western Kenya. Proceedings of the SR-CRSP Kenya Workshop. Small Ruminant Collaborative Research Support Programme, Nairobi, Kenya.

Gatenby R M and Trail J C M. 1982. Small ruminant breed productivity in Africa. International Livestock Centre for Africa, Addis Ababa, Ethiopia.

Getachew Assamenew. 1988. Economic aspects of sheep production in the Ethiopian highland vertisol areas. International Livestock Centre for Africa, Addis Ababa, Ethiopia (mimeo).

Gryseels G. McIntire J and Anderson F M. 1986. Research with a farming systems perspective at ILCA. ILCA Bulletin 25: 17-22.

ILCA 1979. Small ruminant production in the humid tropics. Systems Study No. 3. International Livestock Centre for Africa, Addis Ababa, Ethiopia.

ILCA 1987. ILCA's strategy and long-term plan. International Livestock Centre for Africa, Addis Ababa, Ethiopia.

Mukhebi A W. 1987. Models of economic analysis of animal disease control measures. International Laboratory for Research on Livestock Diseases, Nairobi, Kenya (mimeo).

Ngategize P K and Kaneene J B. 1985. Evaluation of the economic impact of animal disease on production: A review. Veterinary Bulletin 55: 153-162.

Okali C. 1979. Socioeconomic information for ILCA's small ruminant programme in Badeku and Eruwa areas. Oyo State, Nigeria. Humid Zone Programme Working Paper No. 1. International Livestock Centre for Africa, Ibadan, Nigeria.

Raintree J B and Turay F. 1980. Linear programming model of an experimental Leucaena-rice alley cropping system. IITA Research Briefs 1: 5-7.

Sumberg J E and Cassaday K. 1986. Sheep and goats in humid West Africa. International Livestock Centre for Africa, Addis Ababa, Ethiopia.

Sumberg J E, McIntire J. Okali C and Atta-Krah A. 1987. Economic analysis of alley farming with small ruminants. ILCA Bulletin 28: 2-6.

Upton M. 1984. Models of improved production for small ruminants. In: J E Sumberg and K Cassaday (eds.), Sheep and goats in humid West Africa. International Livestock Centre for Africa, Addis Ababa, Ethiopia.

Upton M. 1985. Returns from small ruminant production in South West Nigeria. Agricultural Systems 17: 65-83.

Upton M. 1986. Modelling economic outcomes of livestock production systems. In: N de Ridder, H van Keulen, N G Seligman and P J H Neate (eds.), Modelling of extensive livestock production systems. International, Livestock Centre for Africa, Addis Ababa, Ethiopia.

Verinumbe I, Knipscheer H and Enabor E E. 1984. The economic potential of leguminous tree crops in zero-tillage cropping in Nigeria: A linear programming model. Agroforestry Systems 2: 129-138.

Wilson R T and Bourzat D. 1985. Small ruminants in African agriculture. International Livestock Centre for Africa Addis Ababa, Ethiopia .

Winrock International. 1983. Sheep and goats in developing countries: Their present and potential role. World Bank, Washington DC, USA.

Watish sub-type of Sudan Desert sheep in the Gezira

Previous Page Top of Page Next Page