Earnings and gender wage gap in Lebanon: the role of the human and social capital

ABSTRACT

This paper investigates the earning function and the gender wage gap in Lebanon using human and social capital covariates. The study uses a randomly collected sample from the Lebanese working population in the greater Beirut area applying the counterfactual decomposition and the generalized quantile regressions. Most of the results support previous research which concludes that the human and social capital indicators have a significant impact on earnings and that the wage gap is mainly due to structural effect. Moreover, the results show evidence of quicksand and glass ceiling effects that limit female income.

KEYWORDS:

• Gender wage gap
• human capital
• social capital
• counterfactual decomposition
• generalized quantile regression

JEL CLASSIFICATION:

• J16
• J23
• J24

Acknowledgement

The authors would like to thank an anonymous referee and Iman Khazaal at the Lebanese ministry of labor for helpful comments. The authors would like also to thank the communities of USAL University, LaSagesse University, Kafaat University, LIU University, AUB University, and many others for helping in disseminating the questionnaire used in the study. A special thank you for Dima Hakim for helping in editing.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ To insure randomness, the questionnaire was distributed electronically and physically in different districts in the greater Beirut area. The surveyors distributed the questionnaire randomly to assure as much diversity as possible: poor, medium class, rich, commercial, financial, touristic, industrial, and residential districts were targeted; different markets and shopping places were visited; Christian, Sunni, and Shia districts and other minority groups’ districts were selected. To insure further randomness, surveyors from different backgrounds were recruited: rich and poor, young and old, Christians and Muslims, students and non-students from different social classes and backgrounds.

² Further information about these variables can be obtained from the authors upon request.

³ The assessment of the foreign language proficiency is left to the employee herself (himself) because there is no standardized test to rely upon.

⁴ Results related to both variables are available from the authors upon request.

⁵ Results related to the hourly earnings are available from the authors upon request.

⁶ Generalized quantile regressions (GQR) were conducted in STATA using the genqreg function.