Modeling Supply Response for Saskatchewan Special Crops
Date
1995
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
ORCID
Type
Degree Level
Masters
Abstract
Minogue, L.M., M.SC. University of Saskatchewan. Saskatoon, September, 1994. Modeling Supply Response for Saskatchewan Special Crops.
Supervisors: Dr. R. Gray, and Dr. J. Alston
Special crops are becoming an increasingly important component of Saskatchewan's agricultural sector. Among others, these include lentils, sunflowers, mustard, field peas, canary seed, fababeans, and buckwheat. From the 1982/83 crop year to the 1993/94 crop year, provincial acreage seeded to these five major special crops grew from 430,000 acres to 1,305,000 acres, an increase of 203.5%. Saskatchewan special crop production is a large component of total Western Canadian production. In the 1993/94 crop year, Saskatchewan produced 70.87% of Western Canada's physical special crop output. Since 1985, the province's share of Western Canadian special crop production has grown for all of the major special crops.
Because special crops have become so important to Saskatchewan's agricultural economy, it is important to analyze the economic impacts of programs and policies on these crops. Supply response models are important to these analyses. Several agricultural economists, including Miranda et. al. (1994), Clark and Klein (1992), Meilke (1976), Capel (1968), and Schmitz (1968), have derived provincial or regional supply models for commodities traditionally grown in the prairie provinces but no models have been published for special crops.
The objective of this study is to determine and estimate an acreage response model for Saskatchewan special crops, while reviewing literature related to different types of supply models. In the process of developing these models, two types of supply response models are estimated and compared.
A model suitable for forecasting special crop acreage response was designed by going through a model selection process. The complete systems approach is first considered; a series of supply equations for related outputs are estimated simultaneously, with similar functional form for each equation. Because all parameters are estimated simultaneously, a large number of degrees of freedom are needed. This means it is necessary to make some restrictive assumptions about producer behavior to create enough degrees of freedom to estimate the model. These assumptions took the form of a multistage decision making model.
After this model is estimated and results analyzed, model restrictions are dropped in stages. At each stage of the process, results are estimated and analyzed. The final stage of the process involves a series of ad-hoc models; single equations are estimated independently for each commodity of interest. Each equation includes only variables relevant to the supply of the commodity modeled. Because fewer degrees of freedom are needed, assumptions about producer behavior made for the systems approach are dropped.
For the original systems modes and the final ad-hoc equations, two different functional forms are examined and the most appropriate selected.
In both the first and final steps of this process, translog or log-linear models fit the data better than quadratic or linear forms. For mustard and peas, log-linear ad-hoc equations fit the model best, although all of the ad-hoc equations have several parameters with low t-statistics and incorrect signs. Time trend variables, representing technical change, are the most significant variables in all of the regressions. None of the models
estimated are useful for forecasting provincial special crop supply.
Description
Keywords
Citation
Degree
Master of Science (M.Sc.)
Department
Graduate Studies and Research
Program
Agricultural Economics