REGRESSION EQUATIONS IN ACCOUNT ANALYSIS
It has been customary to express the financial results for a farm in "per acre" terms, the available farms then being grouped according to type and/or intensity of production and their "per acre" results added together and averaged. This is the practice in the Farm Management Handbook of the Universities of Bristol and Exeter where 22 regional types and 14 enterprise types of farm are distinguished. However, the number of farms in each group is never less than 10 and is frequently in excess of 30. Such a rigorous definition of type is not possible in horticulture because of the diversity of holdings and the small number of accounts available. Hence, in horticulture the "per acre" average is a less satisfactory description of the group as a whole.
Horticultural management as one aspect of the horticultural adviser's work first came into prominence in 1960 and it was given further impetus by the introduction of the Small Business Grant Scheme in 1964. A sudden demand for standards by the Advisory Service coupled with the paucity of available financial information on many horticultural holdings led Bristol University to suggest for use in management a method of analysis based on the relationship which was known to exist between costs and output. The Agricultural Economics Department already has sufficient whole-holding accounts for cost-output relationships to be estimated for most types of horticultural holding known to occur in its area. These relationships are most usefully and precisely expressed by fitting regression equations to the data from a sample of holdings which as far as possible are uniform in size, type of production and so on.
These equations are of the form y = mx + c , where y is cost and x is output. The treatment of cost as the dependent variable in a production function is unusual but this approach was deliberate. It was considered that the one reasonably accurate figure likely to be available on all horticultural holdings would be total output, and not the breakdown of cost into its various components. Although y = mx + c is a simple equation, in practice and over the range studied the straight line is a fair approximation of the cost-output relationship of the Bristol data.
Equations of this sort were calculated by G.J. Tyler of the Department of Agricultural Economics and published in 1964 in a brief report called "Cost-Output Relationships in Horticulture". In his report Tyler also