RT Dissertation/Thesis
T1 Smallholder milk production in the Punjab of Pakistan and the evaluation of potential interventions
A1 Teufel,Nils
WP 2007/09/05
AB The potential offered by milk production and marketing to increase rural income has long been recognised. Yet, milk production by smallholder households, the great majority of the 4.1 million households with dairy animals in the Punjab of Pakistan, has remained practically unchanged.
In order to facilitate improved decisions on the allocation of limited development resources, this study analyses smallholder milk production, determines the effects of a variety of technical improvements and evaluates these by their household effects. For this, a linear programming model simulates household decisions based on identified objectives, recorded resources and technical characteristics. Multiple objectives are considered through compromise programming.
Ten improvement scenarios are formulated. These include the introduction of new feeds (maize grain, commercial concentrates, molasses and hybrid sorghum), improved reproductive performance (reduced age at first calving and reduced inter-calving period), improved veterinary health (reduced mortality of calves and of adult females) and the introduction of genetically improved animals, through selective breeding (buffalo and cattle) or crossbreeding (cattle).
Model data were collected during a cross-sectional survey of 322 randomly selected households and a 13-month longitudinal survey of 64 specifically selected smallholders in the central region of the irrigated Punjab of Pakistan.
Cluster analysis identifies seven household classes in the cross-sectional sample. The three largest are considered in the model, characterised as "smallholders with marginal land resources" (class 1), "smallholders with subsistence orientated milk production" (class 2) and "smallholders with market orientated milk production" (class 3). The classification is applied to the longitudinal survey sample by discriminant analysis, identifying 9, 23, and 31 households in classes 1 to 3, respectively.
Herd sizes are similar in household classes 1 and 2, with 2.1 and 2.3 adult female buffaloes respectively, while 3.6 are kept in class 3. Only 0.5 adult female cattle are kept in all classes. Land availability is similar for classes 2 and 3 (2.4 ha), but only 1.0 ha in class 1. Cropping patterns are dominated by wheat and rice. Berseem (Egyptian clover) and sorghum plus maize are important fodder crops.
The nutrient supply calculated from feeding records corresponds well to estimated livestock nutrient requirements. Nevertheless, feed shortage periods in spring and early winter demand attention. Feed energy constrains livestock nutrition throughout the year in all classes.
In addition to maximising family income, households also aim at maintaining their welfare against negative shocks. Also, opportunity costs of family labour are difficult to determine. Thus, maximising farm income and livestock assets (for insurance) as well as minimising loan requirements and family labour use are defined as model objectives in this order of importance. Food requirements are formulated as constraints.
The activity levels produced by the base household model correspond well to longitudinal survey data. However, milk marketing is not only determined by the milk price but also by the reliability of marketing systems. Therefore, the milk price within the model is reduced to simulate recorded milk marketing levels. The price reduction is interpreted as the ?perceived? risk associated with milk marketing at village level.
Model results of the ten improvement scenarios show positive effects for most interventions. Cheap high-energy feeds ("molasses", "hybrid sorghum") provide the greatest benefits for all classes. Increasing annual milk output ("inter-calving period", "selective breeding") is mainly effective in households selling substantial amounts of milk. Reducing replacement requirements ("adult mortality", "age at first calving") does not considerably improve the attractiveness of milk production. High-value concentrates (?maize grain?, ?commercial concentrates?) are unattractive as long as alternatives (e.g. wheat grain) are self-produced at current opportunity costs of land and labour. Finally, reducing calf mortality has practically no effect and crossbred cattle are only attractive under good marketing conditions with effective institutions.
Including only income as objective indicates rising incomes when herd sizes are reduced in favour of cash crops. In a third model version, the effects of higher "perceived" milk prices simulate reactions to improved milk marketing. Despite considerable improvements in the adjusted base model, low-cost feeds still manage to produce considerable benefits.
The study shows that introducing targeted interventions, such as providing cheap high-energy feeds, to smallholder milk production and improving market access are effective approaches for decreasing rural poverty.

K1 Milchproduktion
K1 Haushalt
K1 Lineare Optimierung
K1 Intervention
K1 Bewertung
K1 Pandschab <Pakistan>
K1 Wasserbüffel
PP Hohenheim
PB Kommunikations-, Informations- und Medienzentrum der Universität Hohenheim
UL http://opus.uni-hohenheim.de/volltexte/2007/180