(Non-) discriminatjon in online and fjeld experiments Didier Ruedin University of the Witwatersrand & University of Neuchâtel Workshop on Discriminatjon Neuchâtel, 23 March 2017 image: jazbeck on fmickr
Partjcipants evaluate one job candidate Imagine you are responsible for hiring employees in a large enterprise. You are hiring a receptjonist, with the following requirements : ● focus on customer relatjons ● pleasant and friendly personality with good 0 1 2 3 4 5 6 7 communicatjon skills | | | | | | | | ● dynamic, versatjle, capable to deal with new ● I will invite this candidate for a fjrst informatjon ● resistant to stress, ability to cope with interview. ● I think that I will engage this person. pressure when faced with unexpected ● As an employer, I will enjoy working with this problems ● good management skills and well-organized candidate. ● My co-workers will enjoy working with this You are considering the following CV: ● Ervin Beqiri candidate. ● 11.05.1992 ● My customers will enjoy working with this ● Swiss natjonality candidate. ● mother tongue French ● This person has the right skills for this job. ● good knowledge of English and German ● The profjle of this candidate matches the ● driving licence category B1 requirements of the labour market. ● good IT skills ● By engaging this person, I will take a risk. ● lives in Boudry (NE) ● This candidate will ofuen be on sick leave.
I manipulated skills of the job and ‘origin’ of the candidate Job Skill Name Natjonality receptjonist low François Meylan, Martjn Swiss Baumgartner receptjonist low Dalmat/Ervin Bequiri Swiss director of sales high François Meylan, Martjn Swiss Baumgartner director of sales high Dalmat/Ervin Bequiri Swiss
Manipulatjon check and placebo OK the candidate is a woman the candidate is of Swiss origin the candidate lives in the canton of [Neuchâtel]
No evidence of discriminatjon Dalmat François Ervin François
No difgerence between skill levels Dalmat François Ervin François
meta-analysis In meta-analysis low-skilled jobs have higher discriminatjon rates
In summary, patuerns of discriminatjon are probably equivalent in high/low skilled jobs similar responses in two experiments do I capture antjcipatjon of ‘appropriate’ answers?
Recommend
More recommend
