Migration Analysis and Farm Number Projection Models: A Synthesis

Abstract

Alternative farm number projection models are examined for the Canadian Prairie Provinces including a synthesis of Markov transition probabilities and migration functions. The procedure indicates a potential for overcoming deficiencies of standard farm projection models. Explanatory variables include: oH-farm work, age of operators, capital, and regional economic structure.A CRITICAL analytical requirement for adequate planning for low income problems of farm operators is a model which explains changes over time in farm operator numbers by receipts class. In Manitoba the government has outlined a "stay option" with the policy intent of decreasing the past reductions in farm operator numbers. Clawson expressed the opinion in 1963 that U. S. farm policy has been excessively preoccupied with agricultural supply and demand relations, price supports, surpluses, and related problems [1].Current criticisms of U. S. agricultural policy have made the same point in concluding that the needs of low income farm operators are not being met [20]. Policy planning and evaluation requires an improvement in methods of projecting farm target clientele by receipts class.Past trends illustrate dramatic decreases in farm numbers since the historic peak in the late thirties. The theoretical explanation derived from economic theory and structural analysis of the agricultural industry is straightforward. Given an income inelastic domestic demand, limited export demand for agricultural products, increasing economies of size in farm production, substitution of capital for labor in agriculture, and a high growth rate and high absolute level of wages in the non-farm sectors, decreasing numbers of farm operators are required over time to meet national agricultural production requirements.Farm number projection models have not included relationships which identify the structural significance of the factors outlined. Prior models *Assistance with calculations by G. Gislason is acknowledged, as well as helpful comments by C. F. Framingham, E. W. Tyrchniewicz, and two anonymous reviewers. JAMES A. MACMILLAN is professor and F. L. TUNG and JOHN R. TULLOCH are research assistants in agricultural economics at the University of Manitoba.292 used in projecting farm numbers can be grouped into four categories:1. Age-demographic [1]. 2. Markov chain projection assuming constant transition probabilities [4,10,11,14].3. Requirements based on assumed future average output per farm [2,15].4. Simultaneous analysis of the demand and supply of farm labor [30,32,33]. The age-demographic analysis is the simplest. Entrance and exit of farm operators are projected by age classes over a time period. Markov chain applications are based on the assumption that current values of economic variables depend on the preceding values of the same variables [18] .The requirements approach derives farm operator numbers from projected market-clearing output of agricultural products by means of the projected ratio of average output per farm. Simultaneous demand and supply analysi...

Impact of policy-induced structural change on milk quality: evidence from the Flemish dairy sector

“…The non-stationary approach is useful when there are no direct recorded observations of the transitions, but where these are estimated using additional available economic information (as in Zepeda, 1995a) or when the economic factors that can influence movements have to be filtered out from the transition probabilities [as in HallBerg (1969), Macmillan, Tung & Tulloch (1974) and Zepeda (1995a)]. The present study adopted a stationary Markov process because of its simplicity, the availability of information about the exact number of transitions, and the absence of additional economic information.…”