Propensity score matching without conditional independence assumption - with an application to the gender wage gap in the UK
Type
discussion paper
Date Issued
2003-01-01
Author(s)
Froelich, Markus
Abstract
revised version of Discussion paper 2002-08#### Propensity score matching is frequently used for estimating average treatment effects. Its applicability, however, is not confined to treatment evaluation. In this paper it is shown that propensity score matching does not hinge on a selection on observables assumption and can be used to estimate not only adjusted means but also their distributions, even with non-iid sampling. Propensity score matching is used to analyze the gender wage gap among graduates in the UK. It is found that subject of degree contributes substantially to explaining the gender wage gap, particularly at higher quantiles of the wage distribution. Download Discussion Paper: (pdf, 910 kb)
Funding(s)
Language
English
HSG Classification
contribution to scientific community
Refereed
No
Subject(s)
Division(s)
Eprints ID
15749
File(s)![Thumbnail Image]()
open.access
Name
Froelich_pscore.pdf
Size
909.82 KB
Format
Adobe PDF
Checksum (MD5)
151817fe5037e8f4d1b8e9c9d42ab0e7