Essays on international grain marketing and demand system analysis.

摘要

The objectives of the first essay are to estimate South Korean import demand for grain using both one stage RSDAIDS and two-stage censored RSDAIDS to analyze the effect of tariff elimination as a result of the U.S.-Korea FTA on South Korean import demand for grain. Another purpose of this study is to compare the results of estimates obtained from the two models. Based on prior expectation and empirical results, we conclude that using the Shonkwiler and Yen's estimation method to deal with issues arising from censored time series data provides substantially better estimates compared to that from one-stage RSDAIDS. Most of own price elasticities of South Korean import demand for grains are negative and many of them is elastic. The cross price elasticities within goods indicate that there are significant substitution relationships between the U.S. and Australian wheat, the U.S. and Chinese corn. Expenditure elasticities of Australian wheat, that of ROW corn, and that of Brazilian soybean meal are elastic while that of the U.S. corn are inelastic. As a result of the U.S.-Korea FTA, the import quantity of the U.S. wheat and soybean will be increased about four percent in the South Korean grain market. However, that of the U.S. corn will not be changed.;The purpose of the second article is to elicit South Korean millers' preference and willingness-to-pay for the quality characteristics of hard white wheat that are used in producing all-purpose flour, using two methods (choice-based conjoint analysis and self-explicated approach). Another objective of this study is to compare the predictive validity of these two methods. In specified seven attributes, test weight, moisture, and price significantly affect South Korean millers' utility, but protein contents, ash, dockage, and falling numbers do not when they buy wheat for producing all-purpose flour. South Korean millers are more willing to pay to change the quality characteristics related to the milling yields. To produce all-purpose flour, South Korean millers prefer from 10.5 to 11.5% of protein contents wheat but 12% of high protein contents wheat is not preferred. Along with the research of Srinivasan, we found that the self-explicated approach yields a slightly (but not statistically significantly) higher predictive validity than does the choice-based conjoint analysis.