1 Y 1 Y 0 D = 1 E Y 1 D = 1 E Y 0 D = 1 D = 0 E Y 1 D = 0 E Y 0 D = 0 X 0 1 X = 0 X = 1 Average Treatment Effect ATE 22 E δ =E Y 1 - Y 0 = E Y 1 -E Y

Size: px

Start display at page:

Download "1 Y 1 Y 0 D = 1 E Y 1 D = 1 E Y 0 D = 1 D = 0 E Y 1 D = 0 E Y 0 D = 0 X 0 1 X = 0 X = 1 Average Treatment Effect ATE 22 E δ =E Y 1 - Y 0 = E Y 1 -E Y"

际国
5 years ago
Views:

1 Jan Journal of Xi'an Jiaotong University Social Sciences Vol. 34 Sum No 认为在社会关系发挥正功能的情况下, 网络使用过程中显现的是穆尔 ( Movw) 提及的拥有好的关系的行动者倾向于使用关系的内生性类型 ; 而在社会关系发挥负功能的情况下, 网络使用过程中显现的是有能力的行动者倾向于不使用关系这一个殊异的内生性类型 ; 在此确定基础上, 认为穆尔提出的内生性不具有普适性, 其对网络研究的一般形式的否定是不成立的, 并验证了关系的使用对于收入的效应呈现随着关系使用概率而呈现变化的异质性效应社会资本 ; 同质性 ; 内生性 ; 反事实分析 ; 倾向性评分匹配方法 ; 异质处理效应模型 C A X Mouw sequential search model ex- tensive search model The National Longitudinal Surveys NLS YJA

2 1 Y 1 Y 0 D = 1 E Y 1 D = 1 E Y 0 D = 1 D = 0 E Y 1 D = 0 E Y 0 D = 0 X 0 1 X = 0 X = 1 Average Treatment Effect ATE 22 E δ =E Y 1 - Y 0 = E Y 1 -E Y 0 = πe Y 1 D = π E Y 1 D = 0 - πe Y 0 D = π E Y 0 D = 0 = π E Y 1 D = 1 - E Y 0 D = π E Y 1 D = 0 -E Y 0 D = 0 = πe δ D = π E δ D = 0 1 ATT Average Treatment Effect on the Treated E Y 1 D = 1 -E Y 0 D = 1 E δ D = 1 ATU Average Treatment Effect on the Untreated E Y 1 D = 0 - E Y 0 D = 0 baseline difference E δ D = 0 π treatment pre - treatment heterogeneity ATE π ATT π ATU 1 - π 1 E Y 1 D = 1 -E Y 0 D = = E δ + E Y 0 D = 1 - E Y 0 D = π E δ D = 1 -E δ D = ( 一 ) 反事实分析与倾向性评分匹配方法 E Y 1 D = 1 - E Y 0 D = 0 E δ E Y 0 D = 1 -E Y 0 D = 0 E δ D = 1 - E δ D = 0 1 E δ D = 1 - E δ D = 0 self selection bias ATE 85

3 http / /www. xbskb. cn http / /skxb. xjtu. edu. cn JSNET Propensity Score Matching logist PSM 1 Z 2 Z 3 ATE ATT ATU logit Propensity Score Matching randomized subclassification *** *** ** ATE ATT ATU PSM * * * * * = * *** *** ** = *** PSM Greedy Matching Algorithm *** *** *** *** * ( 二 ) 影响是否使用关系的因素 *** *** *** *** * R

4 3 OLS 1 /10 PSM 2 3 ATT 2 E Y 0 D = 1 - E Y 0 D = 0 ( 四 ) 不同时期使用关系对求职的平均处理效果 ( ATT) ( 三 ) 使用关系对求职的平均处理效果 ( ATT) 4 PSM 16 1 PSM OLS PSM 4 PSM 1 OLS - PSM OLS PSM ATE ATT ATU OLS ATT ATU ATE *** *** *** *** *.996 *.996 * *** *** *** *** *** *** *** *** ATU ATE Bootstrap 50 3 PSM OLS OLS ATU * * *.823 * * * ATU ATE Bootstrap OLS PSM

5 http / /www. xbskb. cn http / /skxb. xjtu. edu. cn OLS PSM 1995 PSM 2001 ignorability assumption pre - treatment heterogeneity PSM heterogeneity of treatment effect ( 一 ) 异质性处理效应模型 heterogeneity of treatment 1995 effect model Brand y i = α + δ i d i + β i X i + u i 1995 d i = 1 = 0 δ ρ α d 0 ρ δ d 0 PSM Brand PSM

6 PSM 1994 ( 二 ) 关系的异质性处理效应分析 1994 PSM Stata HTE Jann B. J. E. Brand Y. Xie Stata module to perform heterogeneous treatment effect analysis. Available from http / /ideas. repec. org /c /boc /bocode / s html 89

7 http / /www. xbskb. cn http / /skxb. xjtu. edu. cn M / / J J J J J M J M / / M JSNET BIAN YANJIE SOON ANG. Guanxi Networks and Job Mobility in China and Singapore J. Social Forces BRAND JENNIE E YU XIE. Who Benefits Most from College Evidence for Negative Selection in Heterogeneous Economic Returns to Higher Education J. American Sociological Review COLEMAN J S. Social capital in the creation of human capital J. American Journal of Sociology Supple- ment GRANOVETTE MARK. The Strength of Weak Ties J. American Journal of Sociology GRANOVETTER MARK. Getting a Job M. Chicago 90

8 IL University of Chicago Press GRANOVETTER MARK. Afterword 1994 Reconsiderations and a New Agenda M / / Getting A Job A Study of Contacts and Careers Second Edition. hicago and London University of Chicago Press MEPHERSON MILLER Lynn Smith - Lovin James Cook. Birds of a Feather Homo - phi ly in Social Networks J. Annual Review of Sociology MORGAN S L HARDING D J. Matching Estimators of Causal Effects Prospects and Pitfalls in Theory and Practice J. Sociological Methods and Research MOUW T. Social Capital and Finding a Job Do Contacts Matter J. American Sociological Review MOUW T. Estimating the Causal Effect of Social Capital a Review of Recent Research J. Annual Review of Sociology NAN LIN. Social Networks and Status Attainment J. Annual Review of Sociology RUBIN D B. Comment on " Randomization Analysis of Experimental Data The Fisher Randomization Test " J. Journal of the American StatisticalAssociation WINSHIP C MORGAN S L. The Estimation of Causal Effects from Observational Data J. Annual Review of Sociology WINSHIP C SOBEL M. Causal Inference in Sociological Studies M / / M Hardy ed. The Handbook of Data A- nalysis. Thousand Oaks CA Sage PARSONS FURUYA S Pal P Kessler. Biogenesis and function of peroxisomes and glycosomes J. Molecular and Biochemical Parasitology Abstract Research on the Problem of Endogeneity of Social Network a discussion on Mouw's views LIANG Yu-cheng Department of sociology and social work Zhongshan University Guangzhou China Based on empirical and theoretical studies this research indicates that under the circumstance where social network performs a positive function individual actors with good relations show a tendency of using a social network which suits the Endogenous Type mentioned by T. Mouw. And when the network develops a negative function those capable actors may tend not to use networks instead which is described in this research as a particular Endogenous Type II. Therefore it is held that it is invalid to negate the universality of endogeneity proposed by T. Mouw and the general form for the study on the network. The study has also confirmed that the effect of the application of relationship on one' s income demonstrates an endogenous effect along with the probability of the application of relationship. Key words social capital homogeneity heterogeneity counterfactual analysis tendentiousness scoring matching method heterogeneity processing effect model

2001 2,, (Devine & Kiefer, 1991), :, ;,,,,?, 300, 57 % ( Granovetter, 1974),, ( Fernan2 dez & Weinberg, 1997) :,, ( Granovetter, 1973),,,,,,,, (Lin, 1

, 4, 1999,, 3 :,,, 1955,,, ;, 1963, 3,,,, 1956 1999 43,?,? 1999, 4,,,, 20 70, ( Granovetter),, 30, (), ( Information Asymmetry), 77 2001 2,, (Devine & Kiefer, 1991), :, ;,,,,?, 300, 57 % ( Granovetter,