1.Introduction
　　Human capital refers to the sum of the stocks of knowledge, physical strength and skills, which are formed and condensed within human being through human resources investment and also are capable of making values added. The smooth transfer of rural labor to urban sectors or non-agricultural industries is restricted by human capital, policies, and institutions and so on. From a long term, with the implementation of various reform measures, the negative impact of external factors tends to be weaker, but the constraint of internal factor (human capital) is difficult to be eliminated in a short term. Therefore, the author attempts to make an investigation on the effect of human capital accumulation on rural labor employment in this paper.
　　
　　2.Literature Review
　　Investigation conducted by Qiren Zhou (1997) on 118 enterprises in Beijing, Shenzhen, Guangzhou, and Wuhan shows that a clear cultural requirement was proposed by 90% of urban enterprises in the recruitment of peasant-workers[1]. Through the investigation and the multinomial Logit model construction of Ling Yu (2002), educational background was an important factor for migrant labor to become management, professional technician or company employees [2]. There are a large number of domestic studies on labor transfer, but most are concentrated in proving that human capital accumulation is conducive to promoting career transition and enhancing employability from a different perspective [3] [4]. Through related literature reviews, it can be seen that most empirical researches are limited to the empirical analyses of education and related objects, and few is on the effect of human capital accumulation on the transfer of rural labor.
　　
　　3.Research Design
　　3.1.Samples Source
　　The data sources from the on-site investigations, which were made by the search group from March 2010 to July 2010 on the human capital, employment and income 2010 of Henan, Jiangxi and Sichuan peasants. Peasants from developed and undeveloped areas of these provinces were selected, respectively, and finally 472 households were collected as valid samples. 1393 labors were totally involved, in which 523 were non-agricultural labors, and 455 were migrant labors). 　3.2.Statistical analysis of relevant index
　　3.2.1 Basic characteristics of migrant labors
　　The majority of rural labors in urban areas are male, higher than 70% in the total, while the female employment rate is less than 30%. Overall, the educational background of migrant labors is low. In this aspect, however, migrant labors still have some advantages compared with the general samples (see Table 1).
　　Table 1: Statistical analysis of relevant indexes of samples
　　
　　
　　3.2.2 Situation of receiving vocational education, training and professional skills and professional skills
　　Investigations show that less than 25% of migrant workers received vocational training. Training contents are concentrated on drivers, electricians, and welders, and followed by decorating, maintenance, and beauty salon. Generally, technical contents take a low proportion in the total training.
　　3.2.3 Employment time in urban areas
　　Some migrant workers get employed in urban areas for at least 6 months, and some have worked there for 40 years. 62% of migrant workers have worked in urban areas for five years. Many migrant workers often return to home and do farm work in busy agricultural season, but go to work in urban areas in slack season.
　　3.3.Model Construction
　　Multinomial Logit model was constructed in this paper. Employment types of migrant labors are shown in Table 2, and the detailed descriptions on all explanatory variables of model are in Table 3.
　　Table 2: Job distribution of migrant labors at different educational backgrounds (Unit: person)
　　
　　
　　Table 3: detailed descriptions on all explanatory variables of model
　　
　　
　　The explanatory variables were expressed with discrete data and had two options or above, and therefore the model in this paper was set in the form of multiple discrete choice model. The probability for migrant labors to select the jth job can be expressed with the following formula.
　　
　　
　　
　　In above equation, i is each migrant labor; k is job type; n is the total number of samples; j is the number of job types; Xi is control variable affecting the job selection of labor. According to the theory of human capital and experience, it can be judged that the expected symbols of all variables are as shown in Table 3.
　　
　　4.Model estimation results and analysis
　　The fifth job type was selected as regression model reference group in this paper, and statistical regression was made on investigated samples with SPSS software. The regression results and tests are shown in Table 4-1, Table 4-2 and Table 4-3, respectively. (责任编辑：南粤论文中心)转贴于南粤论文中心: http://www.nylw.net(代写代发论文_毕业论文带写_广州职称论文代发_广州论文网）