Modelling Agricultural Production Systems using Mathematical Programming

This thesis focuses on the economics of changing farm-level production practices. It is recognised that many decisions made in a firm or a market are interrelated. Economic analyses aiming at predicting behaviour or recommending on alternatives for action need to account for these relations, otherwise the results and prescriptions will be misleading. In each of the four articles a mathematical programming model of an agricultural production system is developed, and empirical analyses are performed using Swedish data. The first article analyses aggregate effects of on-farm potato processing, utilising a partial equilibrium framework. An analysis of the Swedish potato market shows that, in more densely populated regions, on-farm processing is a part of a socially optimal industry structure. Increased import competition results in a larger share of domestic potatoes being processed at the farm-level. In article II, alternative tillage systems for grain production are evaluated. A mathematical programming model with simultaneous selection of crop rotation, machinery investments and scheduling of tillage and drilling operations is developed, utilising discrete stochastic programming to model field time variability. The empirical results show that a tillage system characterised by lower machinery capital and labour requirements may be as profitable as a conventional system. The model from article II is further developed in article III and IV. Article III recognises that policy measures aiming at reducing nitrogen leaching from crop production often have indirect effects. As an example, the empirical results show that subsidies to catch crops and spring ploughing provide an incentive to increase the area of spring crops, such that these subsidies may increase rather than reduce nitrogen emissions. In article III it is also concluded that cost-effective nitrogen abatement requires a mix of various adjustments of production practices, rather than a focus on a few measures. Article IV analyses whether it is necessary to account for environmental and economic risk when analysing measures to reduce nitrogen leaching in crop production. Considering environmental risk increases abatement costs. However, it appears that the benefits from explicitly accounting for nitrogen leaching variability (environmental risk) in the model are rather small, since the majority of the environmental risk is non-diversifiable. The benefits from including economic risk associated with income variability seem