Analysis of future agricultural change : a farm economics approach applied to Dutch arable farming

A. Wossink

    Research output: Thesisinternal PhD, WU


    <p>This study of agricultural change deals simultaneously with: (a) farm planning, ie. the constant adaptation to changing circumstances at the level of the individual farm firm and (b) conditional forecasting, ie. the analysis of alternative agricultural and environmental policy views and their impact.<p><em>Chapter I</em> gives a general introduction and sets out the objectives and scope of the study. The specific research objectives were: (1) to develop a model system based on farm economics to assess the impact at farm and regional level of different scenarios concerning technical developments, agricultural price policies and environmental regulations; (2) to ascertain and describe technical developments and alternative policy options for Dutch arable farming by means of a number of scenarios; (3) to apply the scenarios and <u>part</u> of the system to arable farming in the North East Polder, the region that served as a case study for implementing and testing the system. The North East Polder was selected because of its intensive cropping pattern, because it is a distinct geographical entity and, because of the availability of data in particular. The time horizon was set at 2005 because of the uncertainty of predicting technical and institutional developments over a longer span of time.<p>The attention paid to scenario development and integration of environmental quality aspects in particular, distinguishes the present study from other research on agricultural change at farm level in regions in the Netherlands. Because the analysis included an investigation of the alternative policy views it went beyond the field of regular farm management research. Furthermore it represents a shift from more practice-oriented research towards an exploration of possible and desirable long-term developments including problem perception and problem definition.<p><em>Chapter II</em> reviews and assesses the relevance of a farm economics approach to research on agricultural change. It is contended that in any adjustment in agriculture the family farm is the central decision making unit and that agricultural change comes about by reactions to external forces, of which technical and institutional developments are to be the most influential. Also involved are internal forces related to the technical and financial status of the farm and to behavioural and family-related factors. The Structure-Conduct-Performance framework was used to bring the elements together and draw up model requirements.<p>In <em>Chapter III</em> it is discussed that because of the orientations mentioned at the outset, namely assessing the optimal farm organization for different external conditions (farm planning) and analysis of the effects of policy measures (conditional forecasting) modelling had to combine the normative and the positive approach. The core of the modular system MIMOSA (MIcro MOdelling to Simulate changes in Agriculture), developed for the study, is a single period linear programming (LP) model. Apart from the usual farm activities, the LP model covers an environmental component representing input and leaching of nitrogen and pesticides. The next part of the MIMOSA system combines three modules for additional feedback within and between family farms, not accounted for in the optimization module. The continuation module accounts for changes in the number of entities within each category over time. By means of the innovation adoption module the results of the normative LP model, which indicates the optimal adaptation in farm organization, are finetuned to differences in adaptation behaviour and aggregated to the category level. With regard to feedback between family farms, only land transfer is considered in the MIMOSA system<p>The modular set-up of the MIMOSA system led to the research being divided into three phases: (1) comparative static model calculations for different representative farm types to assess the optimal farm organization for different external conditions and to elucidate the working of environmental economics models; (2) calculations for farm categories to analyse their path of development over time and (3) extension to the aggregated level by a weighted summing of the results from the different farm categories and by accounting for interfarm relationships.<p>In accordance with these phases building the MIMOSA system included: (a) scenario assessment to reduce the different policy views on future price policy and environmental policy, as well as the technical innovations to expect, to a restricted number of consistent, diverging variants; (b) the construction and implementation of an environmental economics model at the farm level; (c) the construction of modules of feedback within family farms to fine-tune the results of the normative linear programming procedure; and (d) the development of an aggregation procedure accounting for regional interdependence between individual farms. Part a, b and c were applied to arable farming in the North East Polder, part d of the MIMOSA system was not implemented in the present study.<p><em>Chapter IV</em> presents the assessment of the scenarios. Variants were operationalized until 2005 for three main fields: technical development; environmental policy regulations; and agricultural price policy measures and general price changes. By combining the variants six scenarios were composed. Comparing the outcomes of Scenarios I and II and of Scenarios IV and V enabled the effects of environmental constraints to be assessed, whereas from Scenarios I and IV and Scenarios II and V the impact of the two price policy variants followed. Scenarios III and VI represented the impact of a compulsory switch to ecological farming. Scenarios I, II and V were considered as the combinations with the greatest practical relevance.<p><em>Chapter V</em> deals with the identification of representative farm types for the population of 864 specialized crop production farms in the North East Polder. Cluster analysis by means of Ward's method was applied to the factor scores from principal components analysis of farm survey data on the 864 entities. This yielded 13 clusters, from which eight representative farm types resulted after combining several clusters according to size (ha) and type of soil without losing essential differences.<p><em>Chapter VI</em> discusses the structure and data use of the environmental economics LP model. An inventory is given of the environmental effects incorporated into the LP model and of the methods used to assess these effects. Later in this chapter the technical innovations of Chapter IV are specified by LP activities. So, for every crop several cropping variants and new crop care methods were defined representing environmentally-friendlier farming techniques. Defining the cropping variants appeared to be timeconsuming because no ready to use technical data were available. Information was collected from many sources and by consulting experts.<p><em>Chapter VII</em> presents the results of LP computations. for the most important farm type in the North East Polder (type IV), representing 239 of the 864 farms in the population. Compared to the basic situation (1989) all scenarios led to dramatic reductions in annual income. For t = 2000 this varied between circa NLG 28 000 for Scenario I, NLG 31000 for Scenario II and NLG 15 000 for Scenario V, for example. Interestingly, in the case of Scenario I pesticide use was reduced by 89 per cent without imposing environmental regulation. This reduction was achieved mainly by technical innovation.<p>An analysis of adaptation, <em>ie.</em> whether and when the optimal LP solutions would be realized by the entities in a specific farm category was added for conditional forecasting. <em>Chapter VIII</em> deals with this analysis of adaptation. Feedback within family farms was implemented and applied to farm category IV. Phase 2 (see above) of the MIMOSA project was executed, in this manner. Aggregation to the regional level was not elaborated; this would have meant an appreciable increase of LP computations and adaptation analyses for the seven other farm types in the North East Polder. Neither was the land transfer module implemented. Simulating the current institutional reallocation rules would require the formulation of two additional representative farm types and the assessment of transitional probabilities for two of the eight initial farm categories.<p>Regarding the application of feedback within family farms to farm category IV, firstly the continuation module accounts for changes in the number of entities by simulating succession. Secondly innovation adoption is simulated. It was assumed that the diffusion of a particular innovation over the entities in a farm category starts as soon as economic advantages result from the LP computations for the farm representative of the category. How rapidly the entities will respond depends on the characteristics of the innovation and on the resistance among the potental adopters. No empirical data were available on this rate of imitation. Instead, parameter estimates for the innovation diffusion model -- the Bass model was selected for this purpose -- were etablished by consulting experts. It appeared that considering farm discontinuation and differences in innovation adoption did not lead to important changes in the implication of the scenarios I, II and V for farm category IV in t 2000.<p>Finally <em>Chapter IX</em> deals with the applicability of the farm economics approach for planning and conditional forecasting, with the most significant results and issues that merit further research. The LP module of the system MIMOSA is an useful instrument for planning ie. to assess the optimal reactions at the farm level to changing conditions. It is a tool to shed light on the interactions of production intensity, environmental aspects and farm income and to compare the implications of policy measures at farm level. Regarding conditional forecasting. the tendencies regarding the implications of the various scenarios are the most interesting outcome. Relationships and trends are more important than the absolute figures, particularly because the application reported covered only one specific group of family farms. It should be noted that the additional insights gained by a farm-based approach (rather than econometric research, for instance) have to be judged in terms of the additional costs involved.<p>Further refining of the LP model should focus on (a) the risks associated with the environmentally-friendlier cropping variants, (b) the incorporation of an organic matter balance, (c) integration of additional aspects of the environmental impact of pesticides, (d) new technical developments and planned policy regulations, (e) adjustment of the organization of the LP input so that farm-specific constraints and price and yield figures can be considered in a more user-friendly way and (f) a more extensive assessment of the representation of family farms with regard to their financial status and with regard to management objectives in relation to the farm family life-cycle.<p>To recap, the major findings and conclusions of the study are:<br/>- A farm-based methodology combining normative and positive analysis can contribute to give insights for both farm planning and conditional forecasting;<br/>- Linear programming of the individual farm is a method wen suited to indicate the trade offs between farm economic aspects and environmental aspects of arable farming for their whole traject of interaction;<br/>- A modular set-up for a farm-based approach has major advantages both with respect to implementation and application. The problem of agricultural change can be studied in an outwardly spiralling manner, firstly at the farm level, and subsequently at the aggregate level;<br/>- A farm based approach is time-consuming and labour-intensive. For conditional forecasting whether the method is preferable to other techniques such as the time series approach, will depend on the specific research question;<br/>- The root causes of the present environmental problems result from failures in the system of economic incentives. In agriculture the economic incentives are largely determined by price and market regulations. As follows from the scenario results, a policy strategy of attuning environmental regulations to price policy regulations is preferable, further in policy implementation attention should be paid to the expected technical innovations;<br/>- Price policies appear to be most important for the future of arable farming, the model results indicate that the targets formulated for the reduction of pesticide use and the emission constraints for pesticides and nitrogen according to the (proposed) Dutch environmental regulations are easily met; ie. with low income losses.
    Original languageEnglish
    QualificationDoctor of Philosophy
    Awarding Institution
    • Renkema, J.A., Promotor
    Award date7 May 1993
    Place of PublicationS.l.
    Print ISBNs9789054850748
    Publication statusPublished - 1993


    • field crops
    • arable farming
    • farm management
    • farm planning
    • linear programming
    • operations research
    • government policy
    • environmental policy
    • environmental legislation
    • air pollution
    • soil pollution
    • water pollution
    • netherlands
    • agricultural policy
    • green revolution
    • economics
    • flevoland
    • future


    Dive into the research topics of 'Analysis of future agricultural change : a farm economics approach applied to Dutch arable farming'. Together they form a unique fingerprint.

    Cite this