commercialized. At that time, labor salary will be determined by the marginal product within agriculture industry.

Thirdly, the full transfer of the agricultural labor force is conducive to fundamentally ease the tension ratio between people and land in agricultural sector, realize the large-scale production and management of agriculture, and make agricultural production's objective shift from output maximizing to profit maximization, and change agriculture organizations from self-sufficiency units into modern enterprises.

Fourthly, the agricultural sector is expected to break the low-efficiency equilibrium of low labor quality and non-skill-biased technique. Based on change of the scarcity of the labor force and the improvement of equipment rates of per capita land, technological advances to save labor and improve work efficiency will be put on the agenda, skill-biased technique to replace non-skill-biased technical will become the main types of technological advances in agriculture, thus upgrading agricultural technology. Fifthly, that agricultural labor force quality improves, agricultural technology upgrades and the agriculture organizations turn to be enterprises will lead to a same result, that is the convergence of return of investment between agricultural and nonagricultural industries, and the most significant is the convergence of labor productivity between these two departments.

Analysis of endogenous human capital accumulation rate is of important theoretical significance. It can eliminate the problem of human capital heterogeneity raised in exogenous analysis and make the heterogeneity come to homogeneity. Then, analysis of agricultural modernization returns to the existing framework of Ranis－Fei model . In the other word, when human capital accumulation rate is endogenous , Ranis－Fei model is still an effective theoretical tool to analyze agricultural development in the conditions of selective labor transfer. The corresponding practical implication is that if there is no discount in the modernization of agriculture in China, its path will not deviate from the full transfer of agricultural surplus labor relying on the industrialization and urbanization, as described in Ranis－Fei model. And its goals are bound to unify a dualistic structure as the model revealed, that is, the general equilibrium or general equilibrium proneness of relative labor productivity in different sectors. Huang’s "blueprint" may simply be a low-level stage goal of agricultural modernization process.