Abstract
Married men earn more than single men, which is a significant finding in labour and family economics. A considerable amount of research literature discusses this estimated wage effect for married men in the United States. This study finds a meaningful impact on wage after controlling for the publication bias with heterogeneity, considering a meta-analysis of 120 estimates. Marriage premium accounts for 6.8% compared to unmarried men with the evidence of publication bias after applying descriptive statistics, funnel graph, simple meta-regression analysis (MRA), and multiple MRAs. Also, this analysis identifies omitted variable bias as another important aspect for explaining this widespread empirical literature.
Highlights
Married men earn more than single men, which is a significant finding in labor and family economics
A systematic and comprehensive meta-analysis of a male marriage wage premium from research literature finds a meaningful impact on wage
It accounts for 6.8 % with evidence of publication bias after applying descriptive statistics, funnel graph, simple meta-regression analysis (MRA), and multiple MRAs of 120 estimated coefficients
Summary
Married men earn more than single men, which is a significant finding in labor and family economics. To be included for meta-analysis, each study was individually reviewed for its empirical estimate of wage premium with criterion’s such as (i) having at least one empirical estimate of the effect of men’s marriage premium.; (ii) how researchers presented their empirical findings because some did not mention independent variables for models using different estimation techniques (some only referred to the premium with marriage status: married, divorced, widowed, and never married; it is hard to identify omitted variables for each model); (iii) standard errors or t-values should report coefficient of marriage premium (iv) studies that are primarily concerned about the marriage-wage premium otherwise, marriage is a common explanatory variable in wage regressions, such as wage differentials, occupational segregation, and other areas in labor economics; (v) regressions included married men only in the United States and not considered women. The STATA command “metatrim” can be used to correct the publication bias, which is used as an iterative algorithm to fill and recompute the mean effect size by adding missing values in the data until the observations are symmetric
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
