The Effect of Major Matching on the Starting Salary of College Graduates: An Empirical Research Based on Propensity Score Matching Method
-
摘要: 高校毕业生就业与所学专业的匹配情况是影响就业质量的重要因素。基于2015年全国高校毕业生就业调查数据,本文考察了毕业生就业与所学专业的匹配情况,以及专业匹配对毕业生工资起薪的影响,并使用倾向得分匹配方法(PSM)对内生性问题进行了纠正。研究发现:有56.1%的毕业生所学专业与工作匹配,有11.7%毫不相关;OLS方法得出专业匹配情况下的工资起薪比专业不匹配高3.4%,使用PSM方法得出专业匹配的工资效应为5%,传统的OLS模型低估了专业匹配的工资效应。进一步的研究显示,相较于“211工程”高校和高职高专院校,专业匹配对非“211工程”本科高校毕业生存在显著的收入效应;与专科和硕士以上学历的毕业生相比,专业匹配对本科学历毕业生具有显著的正向影响,专业匹配比专业不匹配工资高7%左右;经济学类毕业生专业匹配对工资起薪具有更加显著的正向影响,专业匹配比专业不匹配的起薪高15%左右。基于此,高校的专业学科结构设置和人才培养模式应该在专业匹配上做出更加精准的判断,以有效提高大学生的就业质量。Abstract: The match between college graduates’ major and job is an essential factor affecting the quality of employment. Based on the national survey data of colleges graduates’ employment in 2015, this paper investigates the major match of the college graduates, as well as the effects of major match on the starting salary of graduates. It uses the Propensity Score Matching (PSM) method to solve the endogenous problem, and reduce the deviation of results caused by self-selection.The main conclusions and findings are: there are 56.1% of graduates’ majors and jobs match, while 11.7% are irrelevant. Using the OLS method, the starting salary of the major match is 3.4% higher than the major mismatch, and after correcting the endogenous problem by using PSM method, the salary effect of the major match is 5%. OLS regression method underestimates the effect of major match on starting salary. Further research finds that, compared with Project 211 universities and vocational colleges, major match has a significant income effect on non-Project 211 graduates. Also, compared with graduates of college degree and master degree or above, major match has a significant income effect on bachelor degree graduates, the starting salary of the major match is about 7% higher than the major mismatch. Major match of economics graduates has a more significantly positive influence on starting salary, and the major match graduates is about 15% higher than the major mismatch. Based on this, colleges and universities should make more accurate judgment on the specialty matching in order to effectively improve the employment quality of college students.
-
Key words:
- college graduates /
- major match /
- starting salary
-
表 1 专业匹配状况及起薪水平
非常对口 基本对口 有一些关联 毫不相关 1000元以下 5.10% 3.90% 4.50% 6.10% 1001~3000元 38.90% 48.30% 53.80% 54.90% 3001~5000元 31.30% 30.30% 29.60% 27.20% 5001~7000元 11.60% 9.40% 6.50% 6.10% 7001元以上 13.00% 8.00% 5.70% 5.70% 起薪均值 5169.76 4192.39 3598.92 3741.00 总体匹配状况 19.30% 36.80% 32.20% 11.70% 表 2 专业匹配对工资起薪的OLS估计结果
自变量 Ln(Y) 自变量 Ln(Y) 自变量 Ln(Y) 专业匹配 0.0341*
(0.0176)学习成绩中下25% −0.100**
(0.0474)工作单位在省会城市或直辖市 0.131***
(0.0494)非“211工程”本科院校 0.0216
(0.041)学习成绩中上25% −0.120***
(0.0441)工作单位是三资企业 0.112***
(0.0366)“211工程”院校 0.257***
(0.0447)学习成绩前25% −0.039
(0.0443)工作单位是私营企业 −0.0707***
(0.0259)文史哲法、教育学 0.0387
(0.0272)父亲三农类工作人员 0.0184
(0.0322)工作单位是国有企业 −0.0511
(0.0312)经济、管理学 0.0461
(0.0286)父亲普通工作人员 0.0279
(0.0316)工作单位是党政机关及事业单位 −0.157***
(0.0331)理学 −0.0436
(0.0322)父亲专业技术管理人员 0.0163
(0.0285)工作单位属于第二产业 0.0990*
(0.0533)工学 0.011
(0.026)母亲三农类工作人员 −0.0151
(0.0326)工作单位属于第三产业 0.161***
(0.0489)农学 0.0474
(0.039)母亲普通工作人员 0.0311
(0.0296)常数项 7.588***
(0.0837)医学 −0.0809
(0.0524)母亲专业技术管理人员 0.0781**
(0.0319)样本量 5926 R2 0.236 性别 0.0787***
(0.0179)家庭人均年收入20000元以上 0.0738***
(0.0198)本科生 0.234***
(0.0397)工作单位在县级市或县城 0.00279
(0.0513)硕士及以上 0.592***
(0.0454)工作单位在地级市 −0.0181
(0.0513)注:括号内为标准误,*** p<0.01,** p<0.05,* p<0.1 表 3 倾向得分估计结果(logit回归结果)
变量名 系数 P值 变量名 系数 P值 非“211工程”本科院校 −0.004 0.97 母亲为三农类工作人员 −0.114 0.28 “211工程”院校 0.286 0.03 母亲为普通工作人员 0.016 0.88 文史哲法、教育学 −0.222 0.02 母亲为专业技术管理人员 −0.020 0.85 经济、管理学 0.057 0.51 家庭人均年收入20000元以上 −0.072 0.26 理科 −0.173 0.10 工作单位在县级市或县城 0.482 0.01 工学 0.082 0.32 工作单位在地级市 0.441 0.01 农学 −0.635 0.00 工作单位在省会城市或直辖市 0.599 0.00 医学 0.699 0.00 工作单位是三资企业 0.250 0.09 男生 −0.022 0.71 工作单位是私营企业 0.252 0.02 本科生 0.244 0.04 工作单位是国有企业 0.77 0.00 硕士及以上 0.607 0.00 工作单位是党政机关及事业单位 0.721 0.00 学习成绩中下25% 0.208 0.15 工作单位属于第二产业 −0.005 0.97 学习成绩中上25% 0.488 0.00 工作单位属于第三产业 −0.363 0.00 学习成绩前25% 0.850 0.00 常数项 −1.030 0.00 父亲为三农类工作人员 0.109 0.33 父亲为普通工作人员 −0.056 0.60 父亲为专业技术管理人员 0.058 0.56 表 4 不同类型院校、不同学历专业匹配的平均处理效应(ATT)
样本总体 “211工程”高校 非“211工程”本科高校 高职高专 专科生 本科生 硕士及以上 匹配前 0.126***
(0.190)0.072**
(0.031)0.062**
(0.027)0.008***
(0.190)−0.018
(0.032)0.071***
(0.071)0.044
(0.038)最邻近匹配 0.070**
(0.028)0.041
(0.043)0.160
(0.044)−0.001
(0.063)−0.010
(0.055)0.080**
(0.041)0.072
(0.039)半径匹配
(0.01)0.042*
(0.018)0.043
(0.030)0.057*
(0.027)0.046
(0.046)−0.022
(0.036)0.057***
(0.029)0.061
(0.048)半径匹配
(0.001)0.048**
(0.020)0.031
(0.446)0.077**
(0.035)−0.012
(0.056)−0.021
(0.050)0.077**
(0.038)0.057
(0.076)核匹配 0.049**
(0.017)0.038
(0.029)0.062**
(0.027)0.017
(0.044)−0.016
(0.035)0.062***
(0.029)0.046
(0.046)注:括号内为标准误,标准误运用自抽样法(Bootstrap)反复抽样100次得到;*** p<0.01,** p<0.05,* p<0.1。 表 5 不同学科专业匹配状况的平均处理效应(ATT)
文史哲法、教育 经济学 管理学 理学 工学 医学 其他 匹配前 0.053
(0.042)0.299***
(0.048)0.128**
(0.039)−0.002
(0.063)0.178***
(0.378)0.094
(0.115)0.072**
(0.036)最临近匹配 0.021
(0.046)0.170**
(0.076)0.057
(0.049)0.063
(0.119)0.119
(0.053)−0.179
(0.131)−0.019
(0.049)半径匹配
(0.01)0.048
(0.049)0.155***
(0.064)−0.005
(0.414)0.030
(0.078)0.006
(0.038)−0.131
(0.131)−0.014
(0.038)半径匹配
(0.001)−0.018
(0.069)0.131*
(0.074)0.000
(0.058)−0.43
(0.127)0.029
(0.057)−0.101
(0.123)0.005
(0.051)核匹配 0.052
(0.043)0.163***
(0.055)0.005
(0.038)0.040
(0.072)0.004
(0.039)−0.134
(0.132)−0.013
(0.034)注:括号内为标准误,标准误运用自抽样法(Bootstrap)反复抽样100次得到;*** p<0.01,** p<0.05,* p<0.1。 -
[1] 刘扬. (2010). 大学专业与工作匹配研究: 基于大学毕业生就业调查的实证分析. 清华大学教育研究,(6),82—88. doi: 10.3969/j.issn.1001-4519.2010.06.010 [2] 孟大虎, 苏丽锋, 李璐. (2012). 人力资本与大学生的就业实现和就业质量. 人口与经济,(3),19—26. [3] 王亚丰, 李尚群. (2013). 人职匹配的意义与局限. 观察与思考,(1),11—15. [4] 王子成, 杨伟国. (2014). 就业匹配对大学生就业质量的影响效应. 教育与经济,(3),44—52+57. doi: 10.3969/j.issn.1003-4870.2014.03.008 [5] 杨金莲, 张俊涛. (2012). 基于能力倾向的人职匹配理论. 中国成人教育,(13),52—53. [6] 周必彧, 翁杰. (2010). 大学生所学专业与工作岗位的匹配度及其对工资水平的影响. 教育发展研究,(1),87—90. [7] 周丽萍, 马莉萍. (2016). 高校毕业生的就业匹配与工资起薪的关系研究. 教育学术月刊,(4),82—88. [8] Arabsheibani, G. (1989). The Wiles Test Revisited. Economics Letters, (29), 361—364. [9] Bauer, T.K. (2002). Educational mismatch and wages: A panelanalysis. Economics of Education Review, (21), 221—229. [10] Becker, G S. (1964). Human capital: a theoretical and empirical analysis with special reference to education conomics. Chicago: The University of Chicago Press. [11] Buchel, F. (2002). The effects of overeducation on productivity in Germany-the firms’ viewpoint. Economics of Education Review, 21(3), 263—275. doi: 10.1016/S0272-7757(01)00020-6 [12] Buchel, F., Gripd, A., Mertens, A. (2003). Overeducation in Europe: current issues in theory and policy, (3), 65-92. [13] Duncan, G.J., & Hoffman, S.D. (1981). The incidence and wage effect of overeducation. Economicis of Education Review, (1), 75—86. [14] Hersch, J. (1991). Education match and job match. The Review of Economics and Statistics, 73(1), 140—144. doi: 10.2307/2109696 [15] Miller, P.W., & Volker, P.A. (1984). The Screening Hypothesis: an Application of the Wiles Test. Economic Inquiry, (22), 121—127. [16] Nordin, M.I., Persson, R., Dan-Olof. (2010). Education–occupation mismatch: Is there an income penalty?. Economics of Education Review, (29), 1047—1059. [17] Parsons, F. (2009). Choosing a Vocation. Boston: University of Michigan Library. [18] Quintini, G.J. (2011). Over-qualified or under-skilled: A review of existing literature. Working Papers, (121), 12—14. [19] Robst. J. (2007a). Education and Job Match: The Relatedness of College Major and Work. Economics of Education Review, (26), 397—407. [20] Robst, J. (2007b). Education, college major, and job match: Gender differences in reasons formismatch. Education Economics, (15), 159—75. [21] Sattinger, M. (1993). Assignment models of the distribution of earnings. Journal of Economic Literature, (31), 831—880. [22] Schultz, T.W. (1960). Capital Formation by Education. Journal of Political Economy, 68(12), 571—583. [23] Séamus McGuinness. (2006). Overeducation in the labour market. Journal of Econominc SURVEYS, (3), 396—397. [24] Sicherman, N., Galor O.A. (1990). Theory of Career Mobility. Journal of Polotical Economy, (1), 169—192. [25] Verdugo, R R, & Verdugo, N T. (1989). The impact of surplus schooling on earnings: Some additional findings. Journal of Human Resources, (24), 629—643. [26] Velden, R. vander, Smoorenburg, M.S.M.van. (1997). The measurement of overeducation and undereducation: Self-report vs. Job-analyst method. Maastricht: Research Centre for Education and the Labour Market, Maastricht University. [27] Weitzen, S., Lapane, K.L, Toledana, A.Y. et al. (2004). Principles for modeling propensity score in Medical research: a systematic literature review. Pharmacy epidemiology Drug Safe, (13), 841—853.