|
Field
Experimental Design (Public)
|
Before
We design a two-stage hiring game where the two participants (i.e. employers) will see four candidates, and they must decide whether to hire them or not in the first stage. If they decide to hire the candidate, they then need to choose a wage for their hired candidate in stage 2. We will introduce an intermediate hiring competition or a flexible wage scheme in the different treatments. To allow for learning effects, the participants will play this two-stage game in 5 independent rounds. The experiment is based on a between-subject design.
In the baseline treatment (Treatment A0), 2 employers will have a hiring competition if both employers decide to hire the same candidate, and they are free to choose different wage offer to different candidates. As the main experimental variation (Treatment B0), we consider a non-competitive treatment where employers are free to choose different wage offers and always hire the preferred candidate without competition.
In the main treatments, 2 employers are unable to practice wage discrimination, and they can only choose identical wages for all the preferred candidates under competitive (Treatment A1) and non-competitive scenarios (Treatment B1).
The experiment is designed to examine the economic theory which predicts that discrimination can be profitable for employers to segment market and reduce wage payments in the presence of hiring competition (as in Treatment A0), and it is not profitable in the absence of hiring competition (as in Treatment B0). Moreover, the hiring discrimination against minority workers should be greater if wage discrimination is feasible (as in Treatment A0) than if wage discrimination is banned (as in Treatment A1). And such differences should only exist in the competitive environment (A0-A1) and do not exist in the non-competitive environment (B0-B1).
|
After
We design a two-stage hiring game where the two participants (i.e. employers) will see four candidates, and they must decide whether to hire them or not in the first stage (The profile of a candidate includes age, scores of a real effort task and their ethnicity information). If they decide to hire the candidate, they then need to choose a wage for their hired candidate in stage 2. We will introduce an intermediate hiring competition or a flexible wage scheme in the different treatments. To allow for learning effects, the participants will play this two-stage game in 5 independent rounds. The experiment is based on a between-subject design. We will introduce a reservation wage, R0, for each candidate. In the competitive environment (e.g. Treatment A0), one employer needs to choose a wage higher than R0 to successfully hire the preferred candidate and the other employer need to choose a wage higher than 25-R0 to successfully hire the preferred candidate. R0 ranges from [5, ∞]. Therefore, in the baseline treatments, there is an implicit minimum wage of 5 points.
In the baseline treatment (Treatment A0), 2 employers will have a hiring competition if both employers decide to hire the same candidate, and they are free to choose different wage offer to different candidates. As the main experimental variation (Treatment B0), we consider a non-competitive treatment where employers are free to choose different wage offers and always hire the preferred candidate without competition.
In the main treatments, 2 employers are unable to practice wage discrimination, and they can only choose identical wages for all the preferred candidates under competitive (Treatment A1) and non-competitive scenarios (Treatment B1).
The experiment is designed to examine the economic theory which predicts that discrimination can be profitable for employers to segment market and reduce wage payments in the presence of hiring competition (as in Treatment A0), and it is not profitable in the absence of hiring competition (as in Treatment B0). Moreover, the hiring discrimination against minority workers should be greater if wage discrimination is feasible (as in Treatment A0) than if wage discrimination is banned (as in Treatment A1). And such differences should only exist in the competitive environment (A0-A1) and do not exist in the non-competitive environment (B0-B1).
|