USING INPUT DEMAND AND PRODUCTION FUNCTION MODELS TO ASSESS THE NET BENEFITS OF DAIRY HERD IMPROVEMENT

Two methods of evaluating (the net social benefits of the dairy herd-improvement scheme operated by the New South Wales Department of Agriculture are described. The first involves derivation of 'an input demand' function for the herd-recording aspect of the scheme and use of this function to estimate the economic surplus (net of both private and public costs) provided by the service. The second approach involves deriving a production function for milk from which it is possible to estimate the contribution herd-recording and artificial breeding have made to increasing milk yields per cow. Social benefits are shown to have been less than social costs for herd-recording, however dairy farmers have made net private gains. The herd-recording scheme has contained a regressive subsidy element. The production function approach show that artificial breeding and herd-recording were profitable complements in production.     

REGIONAL CONSUMER EXPENDITURE STUDIES USING NATIONAL FOOD SURVEY DATA

Attention has been drawn in N.F.S. reports to the considerable regional variation in the consumption of individual foods. In this paper, the relevant theory and methodology of food expenditure analysis is briefly surveyed and an attempt is made to point out some of the gaps in the currnt state of knowledge. Results of preliminary analysis of regional variations in expenditure on apples are reported and methodological and statistical problems discussed. Finally, suggestions are made as to how some of the problems may be over-come and the lines along which future studies may proceed.     

THE ANALYSIS OF NET FARM INCOME: AN EXAMINATION OF FARM MANAGEMENT SURVEY DATA

The practical significance of the conventional “net farm income” is elusive. It would be useful to have an accepted method of dividing “management and investment income” into its recognised Components—managerial salary and return on tenant's capital. One is a residual if the other can be calculated, and both alternatives are considered. An empirical formula is suggested for estimating the managerial salary which may be imputed to the farmer, taking account of his total turnover, his labour bill and his net farm income. This formula is then applied to Farm Management Survey data, for individual farms and for groups of average, high and low performance (output per £ input). The method may permit closer analysis of relative profitability.     

AN APPLICATION AND EXTENSION OF THE MARKOV CHAIN MODEL TO CEREAL PRODUCTION

Given aggregate time-series data, conventional Markov Chain estimation of the stationary transition probabilities is performed for a model of the cereal market in Great Britain. The possibility of non-stationary probabilities is examined in the light of these initial results, and the basic model developed to allow exogenous variables to enter via their influence on these parameters. The evidence from the paper would appear to suggest that, at least in the case considered, time-varying transition probabilities are a most attractive alternative.     

A STATISTICAL ANALYSIS OF A PILOT SURVEY OF SALT-AFFECTED DAIRY FARMS

A small stratified sample was drawn from irrigated dairy farms judged to exhibit two degrees (high and low) of salting in the soil. Cobb-Douglas production functions were fitted to time series and cross sectional data for each stratum. The results support the conclusion that the data from the two strata can be pooled and that the quality of the soil in this area can be modelled using an analysis of covariance approach. A preliminary confidence interval for the geometric mean of the ratio of the shadow price of water to the price of butterfat for farms in the sample was also calculated. This interval supports the hypothesis that rationed irrigation water is worth more to these farms than the price paid.     

ESTIMATING LINEAR PRODUCTION COEFFICIENTS FROM FARM BUSINESS SURVEY DATA: A NOTE

This paper presents the results of three approaches to estimating linear production coefficients from sample farm data by apportioning total input usage between individual enterprises. It is suggested that the use of Bayesian priors to conduct constrained estimation is preferable to the use of Inequality Restricted Least Squares.     

PRODUCTION RISK AND INPUT USE: PASTORAL ZONE OF EASTERN AUSTRALIA

Few attempts have been made to estimate production functions for the Australian grazing industries. The question of the nature of the effect of input levels on production risk has been broached even more rarely. Previous investigators had to employ models and methods of estimation which embody highly restrictive implicit assumptions about the nature of risk effects. A typical restrictive feature has been the implication that increasing input intensity leads to increasing risk. In this paper, a much less restrictive model and corresponding estimation techniques are brought to bear on individual farm data for 38 properties with 10 continuous years of production records. Perhaps not surprisingly, it is found that some inputs (especially those capital inputs which might normally be thought of as increasing the safety of production) tend to reduce risk.     

A FARM GROWTH MODEL FOR POLICY ANALYSIS IN AN EXTENSIVE PASTORAL PRODUCTION SYSTEM

The development of a simulation model of an extensive pastoral farming system to assist analysts in their assessment of government policy measures is described. The model was designed to simulate, over a number of years, the physical and financial operation of a sheep and beef production system typically found in the North Island hill country of New Zealand. By manipulating model parameters and data related to prices, costs, taxation and credit, a range of policies can be represented and their effects simulated. The model is used to undertake an ex post analysis of the farm-level impact of the supplementary minimum price scheme in New Zealand and to project farm performance following the abolition of the scheme. Consideration is given to the use of the model to represent sheep and beef production systems elsewhere.     

HERD SIZE AND UNIT COSTS OF PRODUCTION IN THE ENGLAND AND WALES DAIRY SECTOR

This paper examines the relationship between herd size and unit costs of production in the England and Wales dairy sector. Following the two-stage estimation procedure of Dawson & Hubbard (1987), longrun average cost (LAC) functions are estimated from cross-section data for five separate years between 1976/7 and 1986/7. Results show that the LAC curves are U-shaped but highly skewed, implying that considerable economies of size are present at low output levels but diseconomies at larger levels are much less marked. Over time, technological change in particular has been shifting the LAC curves downward and to the right and away from the sector's long-run equilibrium. The differential shift of cost curves, depending on the size of production, implies that technological change has been biased in favour of larger herds.     

STRUCTURAL PROJECTIONS OF THE SCOTTISH DAIRY INDUSTRY USING MICRO AND MACRO MARKOV TRANSITION MATRICES

This article analyses changes in the size-structure of the dairy industry in the SMMB area. It compares the performance of projections based on easily available, aggregate data using well-known computer routines with projections made from detailed, infrequently collected, farm-level data. The paper suggests that although yields are continuing to rise this may no longer be sufficient to compensate for the loss of producers, so that there may be a fall in total milk output in the 1980s.     

AN ANALYSIS OF PEASANT FARM PRODUCTION UNDER CONDITIONS OF UNCERTAINTY*

In this paper the impact of cotton on a traditional food crop system in a semi-arid area of Kenya is analysed using a linear programming model and sensitivity analysis. Alternative criteria, maximax, maximin, and a standard average return maximand, are used in turn. The solutions are all evaluated as ‘best year’, ‘worst year’, and average values. This gives a range of results that cakes account of the uncertainty in the farming situation. Several policy conclusions follow from the analysis. It is argued that the approach used is more realistic than an approach in which a single maximand is used, and that the simplicity of the approach brings it within reach of many involved in farm production analysis.     

THE ESTIMATION OF PRODUCTION FRONTIERS: THE AUSTRALIAN LIVESTOCK/CEREALS COMPLEX1

An attempt is made to combine, empirically, the estimation of product transformation surfaces with the more conventional methods of linear supply analysis. This synthetic approach is used to fit simultaneously a system of six aggregate supply functions. The products covered account for more than 70 per cent of the gross value of Australian rural production.     

A SIMULATION MODEL OF AN ARID ZONE SHEEP PROPERTY

A simulation model of a pastoral property in the Western Division of New South Wales is described. Problems of applying the model to analysis of decision situations faced by graziers are reviewed, and the results of some early applications of the model are summarized.     

AN EMPIRICAL INVESTIGATION OF THE HARVEST OPERATION USING SYSTEMS SIMULATION

A simple model of the cereal harvest operation was constructed for the Wimmera region of Victoria. The model was used to investigate factors influential in determining the harvesting costs of a machinery system over a period of years. An evaluation of alternative machinery systems was also conducted using the model. The superiority of systems simulation over static analysis was demonstrated by using the model to incorporate stochastic variables and to handle the dynamic nature of the harvest operation. In addition, the model output was shown to provide more information for decision makers than had resulted from previous Australian studies. It was concluded that systems simulation is an appropriate technique for investigating farm management machinery selection problems.     

A MICRO-ECONOMIC ANALYSIS OF TEA PRODUCTION USING A SEPARABLE RESTRICTED PROFIT FUNCTION

To derive policy-relevant information, the translog form of a separable restricted profit function was used to model empirically estate tea production in Sri Lanka. Data consisted of 87 observations of prices and quantities for various inputs and outputs from 54 different Sri Lankan tea estates. In different tests, the estimated model respected generally the properties implied by underlying assumptions. Results indicated increasing economies of scale in the factory across the sample and optimal field size to be between 300 and 400 hectares. Amongst estimated short-term elasticities, the supply elasticity of made tea was found to be +0.94.     

A COMPARISON OF PROCEDURES FOR ESTIMATING RETURNS TO RESEARCH USING PRODUCTION FUNCTIONS

The production function approach has been one of the two main ex-post procedures used to estimate the rate of return to agricultural research. A critical part of estimating the marginal internal rate of return (MIRR) is the procedure adopted to spread the benefits of research through time. Past studies using this approach have given only brief consideration to this computational procedure. The objective in this study was to review the different computational procedures used and, then, using cross-section production function estimates for U.S. agriculture, determine whether the MIRR estimates are sensitive to the computational procedure used. The results from this comparison indicate a large range in the estimates. The implication, then, is that careful consideration should be given to the choice of computational procedure, both when undertaking such a study and when comparing the results of different studies.     

AN EXAMPLE OF A HEURISTIC LEARNER IN A SIMULATION MODEL

This article deals with the use of a heuristic learner in a simulation model. Simulation modes do not lend themselves to optimizing algorithms thus the use of search procedures are very important when employing a similation model. This paper brings together the concepts of probability and entropy in the use of a learner for simulation models. The learner in this study uses past information as its source upon which to base the next economic decision. Once the decision is made then the outcome of that decision is used to direct the learner for the next decision. This type of learning is consistent and useful for experimental design such as simulation.
Cette article concemc ľusage de ľassimilation intellectuelle heuristique dans une modèle de simulation. Les modèles de simulation ne se prêtent pas a ľoptimation des algorithmes, mais ľusage des mèthodes de recherche sont très importante quand il s'agit ďntiliser des modèles. Ce document rassemble les concepts de probabilité et ďintropée dans ľusage de ľassimilation intellectuelle pour des modèles de simulation. Dans cette étude ľassimilation intellectuelle utilise les informations antécédentes comme référence, et on y base les décisions economiques ulterieurs. Ces décisions prises, les résultats serviront à orienter ľassimilation intellectuelle lors des prochaines décisions. Ce type ďassimilation intellectuelle est logique et utile au model expérimental comma celui de la simulation.     

AN ECONOMIC RESPONSE MODEL OF HERBICIDE APPLICATION FOR WEED CONTROL

A biologically realistic model of crop yield response to herbicide application is presented. It includes functions for weed mortality from herbicide application and yield loss due to surviving weeds. The optimal herbicide rate and two types of decision thresholds are derived theoretically and illustrated with empirical examples. Responses of the various decision criteria to changes in parameters are also examined theoretically and empirically. A multidimensional threshold for weeds based on weed density and weed-free yield is presented. The issue of farmers using other than officially recommended herbicide rates is discussed.