Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/35658 
Year of Publication: 
2009
Series/Report no.: 
IZA Discussion Papers No. 3972
Publisher: 
Institute for the Study of Labor (IZA), Bonn
Abstract: 
Today there is a variation within the EU to what extent nations allow for situation test results to constitute mass of evidence in court in order to prevent ethnic discrimination. In the UK The Equality and Human Rights Commission has the right to conduct discrimination tests and to even prosecute firms, implying that discriminating firms face the risk of a significant penalty. Other European countries have been reluctant to use such tests as a tool for counteracting discrimination and discuss a much softer version with only monitoring. In this study two labor market field experiments, sending qualitatively identical job applications with randomly assigned Swedish and Middle Eastern sounding names to employers, show that ethnic discrimination exists in hiring in the Swedish labor market. In both studies extensive media coverage occurred when being only halfway finished informing employers of their hiring practices being monitored by such situation testing. This study utilizes these unique events and the data from the experiments to perform a difference-in-differences analysis of whether discrimination decreased after the media coverage. The results reveal no sign of employers changing their hiring practices when being aware of running the risk of being included in such an experiment. This suggests that the detection risk alone is not sufficient if authorities wish to use field experiments as a discrimination prevention strategy. Instead, it must be combined with some penalty to become effective.
Subjects: 
Ethnic discrimination
correspondence testing
situation testing
field experiments
JEL: 
J64
J71
Persistent Identifier of the first edition: 
Document Type: 
Working Paper

Files in This Item:
File
Size
177.85 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.