Efficiency of scottish farms: a stochastic cost frontier analysis

Abstract In this paper the relative cost efficiency of Scottish farms is determined, and variables that explain this efficiency by farm type are identified and implications discussed. A panel dataset from the Farm Accounts Scheme (FAS) survey for the period 1997-2004 was used for the estimation. A cost efficiency indicator was measured using a fixed effect panel data regression. Further analysis, to explain the efficiency results, indicated the presence of important farm size and regional effects. However, other variables, whilst statistically significant, did not produce a consistent effect across the different farm types. Keywords: Stochastic cost frontier analysis, cost efficiency, Scottish farms, Common Agricultural Policy JEL Classification: Q12 (ProQuest: ... denotes formulae omitted.) Introduction For almost half a century, the European Union's (EU) Common Agricultural Policy (CAP) supported increases in farm production with great success. Eventually, however, unacceptable levels of overproduction, budget cost pressures, accusations of excessive market protection and distortion, and expansion of the EU, along with concerns about the environmental impact of agricultural intensification, all contributed to growing support for fundamental reform of the CAP. The new agricultural policy measures adopted by EU farm ministers in 2003, seek to reform the CAP in ways that will enable EU farmers and their businesses to become more market orientated. One of the possible areas of adjustment is related to input use. This poses two questions that this paper tries to answer: first, how heterogeneous are Scottish farms in terms of their efficiency with respect to input use (i.e., cost efficiency) and second, if such heterogeneity exists, what are its causes? Most farm efficiency estimates for the UK are for England and Wales (Dawson, 1985; Wilson et al., 1998; Wilson et al. 2001; Thirtle and Holding, 2003; Hadley, 2006). In the case of Scottish agriculture, Santarossa (2003) and Barnes (2005) used stochastic production frontiers to investigate efficiency using similar explanatory variables, but producing different results regarding farm size. The purpose of this paper is to construct efficiency indicators for Scottish agriculture by farm type using a stochastic cost frontier approach that recognises the multi-output nature of farming, and offers certain advantages over the stochastic production frontier approach.2 The paper starts by outlining the dataset used for the analysis and then describes the cost frontier estimation procedure. Thereafter, the cost efficiency results by farm type are presented, followed by an analysis of their determinants. The paper ends with a set of concluding remarks. Measurement of Relative Cost Efficiency in Scotland This section starts by presenting the data available to estimate the cost efficiency indices and then sets out the methodology used for computing these indices. Data used for the estimation The data used in the analysis are from Scotland's Farm Accounts Scheme (FAS), which annually records financial and non-financial data for a selection of full-time farms across Scotland. The dataset covers the eight year period of 1 997/98 to 2004/5 (i.e., the crop years of 1997 to 2004). The criteria used to select the farms were that they should be present in the 2004/05 survey, and also that they were in the sample for at least five years. This resulted in an unbalanced panel dataset of 358 individual farms. Table 1 summarises this sample by farm types and their respective main outputs. Costs and outputs by farm type were computed directly from the FAS data. Costs were allocated to one of four groups: materials (e.g., feed, fertiliser); purchased services (e.g., contract work, crop protection costs); labour (e.g., all labour used including that of the farmer, farm family, business partners and hired workers); and capital (e. …