FORECASTING MILK OUTPUT IN ENGLAND AND WALES*

This paper evaluates the ablility of an autoregressive integrated moving average (arima) model to forecast monthly milk output before and after the imposition of quotas in April 1984. In the post quota period the standard arima model is modified to a ‘discontinuity’ model allowing the quota to be represented explicitly. It is concluded that for a medium term forecast this is preferable to the standard arima model, whilst for one period ahead forecasts the models are equally good.     

A FORECAST OF MILK SUPPLY IN ENGLAND AND WALES

This paper reports the problems encountered in applying Markov Chain analysis to the Permanent sample of milk producers in England and Wales, and presents predictions obtained for producer numbers and milk output, by regions, for the years 1970/71 and 1975/76. As a consequence of the “pool of entrants” problem, dificulties arise both in estimating the required transition probabilities and in forecasting producer numbers. Solutions to these, and to other problems concerned with obtaining prediction of milk output, from those of producer numbers, are proposed. It is forecast that producer numbers will decline by 18 per cent and 33 per cent by 1970/71 and 1975/76 respectively, with the largest proportional decline occurring among smaller producers. Two forecasts are made of milk supply and these are compared with estimates made by the Milk Morltetiiic Board.     

PRODUCTION CHOICE IN THE ENGLAND AND WALES DAIRY SECTOR

This paper seeks to develop the understanding of the ways in which factors other than price influence the production decisions of dairy farmers. A theoretical model based on a profit function is used to emphasise the way in which changes in the farm's technological environment will influence its production choices. The theoretical model is used as the basis of an empirical study of production choice in the England and Wales dairy sector. Elasticities and factor biases are reported.     

GROWTH AND SIZE ECONOMIES OVER SPACE AND TIME: WHEAT-SHEEP FARMS IN NEW SOUTH WALES

Time series and cross-sectional data for 18 years on 26 N.S.W. wheat-sheep properties are combined to test whether farms that grew in size over time reaped the postulated size economies derived from the cross-sectional statistical cost functions. Realised economies of farms that remained in the industry and changed in size during the 18-year period were found to be significantly greater than measured potential economies from 'static' cost functions would have suggested. The 'dynamic economies of growth' differed significantly amongst farms indicating there was no single economy of size curve along which farms moved.     

MANAGEMENT AND SIZE ECONOMIES IN THE ENGLAND AND WALES DAIRY SECTOR

Analyses of size economies in the England and Wales dairy sector have generally been made on the basis of comparisons of input-output measures. These measures have been classified according to input use and farm size and have led to conclusions about efficiency and structural change. This paper examines the influence of managerial ability on economies of size using econometrically estimated long-run average cost (LAC) functions from Milk Marketing Board data for 1980/1. The results show that the LAC curve is U-shaped though skewed to exhibit greater economies than diseconomies of size. Better managed farms are shown to produce any given level of output at lower average cost. Moreover, they have larger optimal levels of output.     

TECHNOLOGICAL CHANGE IN THE ARID ZONE OF NEW SOUTH WALES

The results described challenge the generally accepted interpretation of the factors underlying the changes in sheep numbers in the arid zone of New South Wales. The extent to which long-run changes in rainfall have been ignored is highlighted. Estimates of output per man are derived, using the C.E.S. production function, which show that there has been a small positive rate of technological change. Three-quarters of such technological advance has been due to factors which affect numbers of sheep carried, the remainder due to factors affecting wool production per sheep.     

QUARTERLY ESTIMATES OF THE DEMAND AND PRICE STRUCTURE FOR MEAT IN NEW SOUTH WALES

Multiple regression estimates of demand and price relationships for fresh beef, lamb and mutton in the N.S.W. livestock auction, wholesale, and retail markets during the period from January 1951 to June 1963 are presented. The results show that direct price elasticities of demand were negative, and of greatest absolute value in the retail market. Mutton is shown to have been a close substitute for beef and lamb, but the latter were not close substitutes with respect to price.     

COMPARATIVE FORECAST ACCURACY IN THE NEW SOUTH WALES PRIME LAMB MARKET

The market for Australian prime lamb is characterised by high production seasonality and a highly competitive retail demand. Because these factors often translate into substantial market variability, regular forecasts of supply and demand are important requirements of lamb market participants. There has been some forecasting activity in the state and national lamb markets but it has been a somewhat controversial activity. This paper assesses the comparative forecast accuracy of a range of methods in the New South Wales lamb market. The results indicate that no single method is clearly superior in all situations and the greatest scope for improving forecast accuracy in the New South Wales lamb market is through the use of combined econometric and naive approaches.     

THE EVOLUTION OF AGRICULTURAL EXTENSION IN ENGLAND AND WALES PRESIDENTIAL ADDRESS

This paper traces the evolution of agricultural extension in England and Wales and reflects on the reasons for the changes which have occurred in its structure, direction and funding. It describes the change from an extension to a consultancy organisation, analyses the effects of such a change, and examines some of the early mistakes in undertaking it. Questions are raised about whether advice paid for by the client improves the focus of extension and enhances the effectiveness of the relationship between adviser and client. It also questions whether Government policies towards agriculture could be equally well carried out by contracting out the work of extension to the private sector.     

LAND TENURE IN THE CYPRESS PINE AREAS OF NORTH-WESTERN NEW SOUTH WALES

In New South Wales there are one million acres of land bearing cypress pine and held under lease from the government. This land is suitable for the joint production of timber and livestock or for timber production alone. Alternatively, the timber can be harvested to provide capital for property development. Farm management plans have, however, been constrained by the land tenure policy which aims to promote both closer settlement and timber production. In this paper the opportunity costs of the current tenure system are evaluated against the freehold situation. The conditions which favour either specialized timber production or the joint production of timber and livestock are examined. Finally, the cost of achieving a timber production target is examined in terms of loss of income to the individual grazier and the nation.     

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.     

MARKETING MARGINS IN THE MEAT SECTOR,ENGLAND AND WALES 1978–1987

The paper presents a basic analysis of price spread data for beef, lamb and pork over the period 1978 to 1987. The results are reviewed in the light of existing margin theory in order to assess how far the characteristics of price and margin behaviour today differ from those highlighted in earlier work. It is concluded that while short run behaviour is well explained, further research is required if we are to fully understand the factors influencing the longer term development of margins in the meat sector.     

LONG-RUN AVERAGE COST CURVES IN THE ENGLAND AND WALES DAIRY INDUSTRY - REPLY

Burton, Ozanne and Collinson (1993) present an interesting comment on the issue of long-run average cost (LAC) in the dairy industry. Since their results essentially substantiate the conclusions of Hubbard (1993), the following points are offered as further thoughts which may be of relevance in any subsequent analysis and estimation.     

CHANGES IN CEREAL PRODUCTION AND YIELD VARIABILITY ON FARMS IN SOUTH EAST ENGLAND

Data from 16 wheat growers and 18 barley growers in South East England were investigated for changes in variability of production and yield between the periods 1964-74 and 1975-84. Sixteen per cent of the increase in variability of wheat production was due to changes in yield variances and covariances. The increase in variability of barley production was not statistically significant. Variability of yield increased for both wheat and barley. Wheat yields became more positively correlated between farms, but barley yields did not. The results for wheat were consistent with the hypothesis of an increase in the number of controlled factors in the production process. Particularly important was thought to be the use of fungicides in the control of many cereal diseases. Increased heterogeneity of the barley crop may have contributed to the decline in the number of between-farm correlations for that crop.     

AN APPLICATION OF MARKOV CHAINS TO FARM TYPE STRUCTURAL DATA IN ENGLAND AND WALES

Markov Chain methods of projection are applied to farm type structural data derived from the returns in the Annual June Agricultural Census in England and Wales. Three sets of projections, using different assumptions, are made of the numbers of holdings by farm type in 1969 and of the movement of holdings between farm type during 1968/69. These are compared with the actual Census returns for 1969. Use of a four-year average transition probability matrix gives the best overall results on the basis of that comparison and this matirx was then used to project farm type structure in England and Wales for 1971 and 1973. These projections indicate that between June 1969 and June 1973 the number of all holdings with 275 or more standard man days (SMD) will decline; those classified as Mixed, Livestock, and Pigs and Poultry are likely to show the largest proportional declines. Some 50 per cent of holdings are likely to be below the 275 SMD limit. Since the assumptions made may not be fulfilled in every respect, it is unlikely that the outcome in 1971 and 1973 will be precisely as predicted, but the pattern and direction of the projections are likely to hold.     

DISTRIBUTIONS OF INDEMNITIES FOR CROP-INSURANCE PLANS: WITH APPLICATION TO GRAIN CROPS IN NEW SOUTH WALES

A review of crop-insurance schemes is followed by a discussion of a guaranteed-yield, crop-insurance plan. General formulae for the distribution function and mathematical expectation of indemnities for the insurance plan are presented in terms of the distribution of crop yields. Three special cases are considered in which the original yields, the square root of yields, and the logarithm of yields are normally distributed. The insurance plan is applied on a regional basis for wheat and sorghum production in N.S.W. Given distributional information on the crops obtained from a simulation model, expected indemnities are calculated for four different insurance plans.