A diagnostic system to assess sustainability at a farm level: The SOSTARE model

The paper presents a model for integrated sustainability assessment at a farm scale. The SOSTARE model (analysis of farm technical efficiency and impacts on environmental and economic sustainability), in fact, aims at providing a diagnostic tool to farmers and advisory services to assess the general performance of the farm, explore in detail any perceived weaknesses in farm management and to investigate the impact of changes that might improve efficiency. The model is derived from a survey of 68 farms. The sampled farms have been selected in order to represent the panorama of farm typologies that are characteristic of the agriculture in one of the most intensively cultivated areas of Europe, the Po plain in Northern Italy, and include multi-functional and non-multi-functional farms, low-input and conventional management typologies, subdivided into production of rice, cereals, and livestock (meat and milk). The SOSTARE model is based on a set of indicators, which are aggregated in a stepwise fashion to provide the user with an immediate valuation of a farm's performance. The aggregation framework was structured on the following steps: statistical analysis of all the collected input data and indicators to identify relevant variables and reduce redundancy; identification of the range of variation of the selected basic indicators and normalisation; definition of weights through statistical analysis and expert judgement, aggregation. The result is a synthetic indicator for each of the three analysed aspects of sustainability (economic, agronomic and ecological), which are decomposed into a total of twelve sub-dimensions. These include: fertilisers, water and agrochemical management; cropping system; energy input; value of production; value added; farm household income; CAP independence; farm business diversification; natural value of the farm and functional landscape pattern. In practical terms the output of the model is provided to the user as original indicator values in the form of tables, bar diagrams corresponding to each aggregation step and a radar diagram representing the overall farm performance in the twelve sub-dimensions. Furthermore, results can be related either to the optimal performance or to a reference situation (i.e. the average of the farm typology, the average of all farms). Results show that the model is able to react to different types of farm management that were studied in the reference sample of farms and that weaknesses of farm performances can be clearly identified. Moreover, its application over time would provide information that could be used to assess and monitor policy performances at a regional level.