Gender Wage Gap Accounting: The Role of Selection Bias

Abstract

Mulligan and Rubinstein (2008) (MR) argued that changing selection of working females on unobservable characteristics, from negative in the 1970s to positive in the 1990s, accounted for nearly the entire closing of the gender wage gap. We argue that their female wage equation estimates are inconsistent. Correcting this error substantially weakens the role of the rising selection bias (39 % versus 78 %) and strengthens the contribution of declining discrimination (42 % versus 7 %). Our findings resonate better with related literature. We also explain why our finding of positive selection in the 1970s provides additional support for MR's main hypothesis that an exogenous rise in the market value of unobservable characteristics contributed to the closing of the gender gap.

Unobserved selection heterogeneity and the gender wage gap

“…The choice of instrument follows the previous literature, and has the advantage of making this study directly comparable to the others that have relied on this same instrument (Bar et al, 2015;Heckman, 1974;Mulligan & Rubinstein, 2008). The choice of instrument follows the previous literature, and has the advantage of making this study directly comparable to the others that have relied on this same instrument (Bar et al, 2015;Heckman, 1974;Mulligan & Rubinstein, 2008).…”

“…Vella, 1998). In particular, Bar et al (2015) show that a generalized version of the standard selection model uncovers a much smaller role of selection bias than what was documented by Mulligan and Rubinstein (2008). Mulligan and Rubinstein (2008) rely on a parametric correction method for measuring the gender wage gap in the USA, even though extensive literature has demonstrated the sensitivity of selection models to several of their modeling assumptions.…”

“…In contrast to a parametric selection correction, the bounding approach relying on an instrument imposes fewer restrictions as it need not model selection. This paper is mostly related to the approaches that allow selection to depend on unobservables and rely on an exclusion restriction, such as the approaches in Mulligan and Rubinstein (2008), Bar et al (2015), Blundell et al (2007), Huber and Melly (2015), and Arellano and Bonhomme (2017). In Blundell et al (2007), the availability of an instrument 6 is combined with a positive selection assumption and an additivity restriction in the wage equation for an empirical assessment of the gender wage gap in the UK.…”

“…Blau and Kahn (2006) have found women's wage gains in the USA to be lower than the observed wage gain of women in the labor market. Using an alternative parametric selection model, Bar, Kim, and Leukhina (2015) have found that selection accounts for a modest part of the convergence, and that reduced discrimination is an important piece of the story. 2 Using a parametric selection correction and identification-at-infinity methods, Mulligan and Rubinstein (2008) have argued that the entire convergence is due to selection and changes in the pool of women that enter the labor market.…”

“…First, it allows for comparability with the existing literature, as its use still remains in recent studies of the gender wage gap (Bar et al, 2015;Mulligan & Rubinstein, 2008). The advantage of this instrument is twofold.…”