The risk-return trade-off to diversified agriculture in Malawi : a quadratic programming approach

Thesis (MCom)--Stellenbosch University, 2021.ENGLISH SUMMARY : The rapid growth in the human population has sparked shifts in the way agricultural sectors have evolved in different countries. Most developed countries and those with a commercially driven agricultural sector have placed emphasis on increasing productivity in a bid to ‘produce more for less’. Developing countries, by contrast, are often dominated by subsistence agriculture where the focus lies on ensuring household food security rather than profit maximisation. Malawi falls into this category with a vast majority of the working population involved in agriculture – more specifically, in the smallholder sub-sector. The risk of crop losses in these countries has dire consequences for people reliant on these crops for their everyday meal. Minimizing such risk in countries like Malawi is therefore of paramount importance. Many studies, such as the one conducted by Ibrahim (2015), place diversification at the heart of risk management within the agricultural context. Consequently, this study investigated the use of diversification as a tool to minimise the levels of risk faced by smallholder farmers in Malawi. Studies by Mango, Makate, Mapemba and Sopo (2018) and Kankwamba, Mapila and Pauw (2013) analysed the determinants of diversification in Malawian agriculture and the current levels of diversification within the country’s agricultural sector. Their results provided insight into the factors influencing diversification and indicated a bimodal distribution for the number of crops grown – peaking at three as well as one. Evidently, the importance of diversification has already reached Malawian smallholder farmers. However, minimal research has been done into the optimum diversification strategies for these farmers to implement on the smallholder level. Some success optimising cropping portfolios for smallholder farmers in Malawi was found using Quadratic Risk Programming. However, that particular research called for an updated and more data accurate investigation. Accordingly, this study implemented the Quadratic Risk Programming model on a large sample of smallholder farmers in the southern region of Malawi. Six models were created, varying the size of the smallholder field and the capacity of the farmer to apply inorganic fertiliser. Five primary crops, namely maize, soybeans, groundnuts, common beans and sweet potatoes, were identified and their performance was analysed over three consecutive years. Each model included a variance-covariance matrix, incorporating the relationships between crops to derive optimized cropping portfolios according to the desired level of risk exposure. For small farms, the results showed that, of the available 2 acres, 1,3 acres should be allocated to maize and the balance shared between groundnuts and beans. A ratio favouring beans gave lower risk than when groundnuts were favoured. However, models for medium and large farms recommended an average allocation of 50 percent of their arable land to groundnut production. In consideration of food security, all models contained a minimum threshold for maize growth. The results for all fertilised farm models indicated sweet potato growth at the maximum constraint, prompting the recommendation for improved storage and marketing facilities for this crop in Malawi. Finally, recommendations were made regarding the use of the state-owned marketing platform, ADMARC, to protect farmgate prices and stimulate an agricultural environment conducive to the findings of this thesis.AFRIKAANSE OPSOMMING : Die vinnige groei van die menslike bevolking het na verskuiwings gelei in die manier waarop landbousektore in verskillende lande ontwikkel het. Die meerderheid ontwikkelde lande en dié lande met ’n kommersieel gedrewe landbousektor benadruk verhoogde produktiwiteit in ’n poging om ‘meer vir minder te produseer’. Ontwikkelende lande, in kontras, word in baie gevalle gedomineer deur bestaanslandbou, waarin die fokus val op die versekering van die huishouding se voedselsekuriteit eerder as winsmaksimering. Malawi val binne hierdie kategorie, met 85% van die werkende bevolking wat in landbou betrokke is – meer spesifiek in die kleinboersektor (Drope, Makoka, Lencucha & Appau, 2016). Die risiko van gewasverliese in hierdie lande hou ernstige gevolge in vir mense wat op hierdie gewasse staat maak vir hulle daaglikse kos. Die minimalisering van sulke risiko’s in lande soos Malawi is dus baie belangrik. Baie studies, soos die een van Ibrahim (2015), plaas diversifikasie sentraal tot risikobestuur in die landboukonteks. Gevolglik het hierdie studie ondersoek ingestel na die gebruik van diversifikasie as gereedskap om die vlakke van risiko wat deur kleinboere in Malawi ervaar word, te minimaliseer. Studies deur Mango, Makate, Mapemba and Sopo (2018) and Kankwamba, Mapila and Pauw (2013) het die determinante van diversifikasie in die Malawiese landbou en die huidige vlakke van diversifikasie in die land se landbousektor geanaliseer. Hulle resultate verskaf insig in die faktore wat diversifikasie beïnvloed en het ’n bimodale verspreiding vir die aantal gewasse wat gekweek is, getoon – wat op drie sowel as een ’n hoogtepunt bereik het. Klaarblyklik is Malawiese kleinboere reeds bewus van die belangrikheid van diversifikasie. Daar is egter minimale navorsing oor die optimum diversifikasiestrategieë vir hierdie boere om op die kleinboervlak te implementeer. Msusa (2007) het ’n mate van sukses behaal met die optimalisering van oesportefeuljes vir kleinboere in Malawi deur gebruik te maak van Kwadratiese Risikoprogrammering (Quadratic Risk Programming). Sy navorsing het egter die behoefte aan ’n opgedateerde en akkurater data-ondersoek uitgewys. Gevolglik het die huidige studie die Kwadratiese Risikoprogrammeringsmodel op ’n groot monster kleinboere in die suidelike streek van Malawi geïmplementeer. Ses modelle is geskep, met verskille in die grootte van die kleinboere se grond en die vermoë van die boer om anorganiese bemesting toe te dien. Vyf primêre gewasse, naamlik mielies, sojabone, grondbone, gewone bone en patats, is geïdentifiseer en hulle prestasie is oor drie opeenvolgende jare geanaliseer. Elke model het ’n variansie-kovariansie matriks ingesluit wat die verhoudings tussen gewasse geïnkorporeer het om geoptimaliseerde oesportefeuljes volgens die gewenste vlak van risikoblootstelling af te lei. Vir klein plase het die resultate getoon dat 1,3 hektaar van die beskikbare 2 hektaar aan mielies toegeken moet word en dat die balans tussen grondbone en bone verdeel moet word. ’n Verhouding ten gunste van bone het gelei tot minder risiko in vergelyking met grondbone. Modelle vir medium-grootte en groot plase het egter ’n gemiddelde toekenning van 50 persent van hulle bewerkbare grond vir grondboonproduksie aanbeveel. Met betrekking tot voedselsekuriteit het al die modelle ’n minimum drempel vir mielies bevat. Die resultate vir al die bemesde plaasmodelle het patatgroei teen die maksimum beperking aangedui, wat gelei het tot die aanbeveling dat verbeterde stoor- en bemarkingsfasiliteite vir hierdie gewas in Malawi verskaf moet word. Laastens is aanbevelings gemaak oor die gebruik van ADMARC, die bemarkingsplatform in staatsbesit, om plaashekpryse te beskerm en ’n landbou-omgewing te stimuleer wat bevorderlik is vir die bevindinge van hierdie tesis.Master